Update to OTW Signal, May 2023

A few days ago we ran an article with an excerpt from an interview with a member of our Legal Committee. That article featured the opinion of one of our 900+ volunteers. It does not represent an official position on the part of the OTW or its Board of Directors. We sincerely apologize for the hurt and confusion we have caused, and we have removed the excerpt.

As fan work creators and users of AO3 ourselves, we understand our users’ concerns around this issue and are taking these very seriously.

2023-05-12 Retired Personnel
Uncategorized
  1. M commented: Not the best holding statement. Curious to know whether OTW has any corporate comms professionals on staff? Pop up a role and I’m sure one of us can lend a hand at this point. 🤦
    2. Reply
    • OS commented: it's funny you ask because this was literally written by the OTW Communications Committee chair, who is a corporate communications professional.
      • Reply
      • M commented: When I look them up they appear to be a research manager at an art faculty but alright. I’ll take your word for it and amend my comment to say: if so, it’s tragic that this was written by a corp comms professional and it reflects badly on them and the board.
        • Reply
  2. Vaynglory commented: "the opinion of one of our 900+ volunteers" - you mean the chair of the OTW's legal committee? Deeply concerning that someone in such a position is so enthused about AI scraping the Archive's works for corporations to profit from. This non-apology seems more like an attempt to sweep the issue under the rug than anything.
    3. Reply
    • depizan commented: Exactly. Don't hide that Rosenblatt is the legal chair. She's not just a randomly picked volunteer, she's someone with a position in the organization that's related to the topic she was discussing. Those of us who use (and donate) to AO3/OTW have good reason to be concerned by her words. This is not the reassurance you (AO3/OTW) seem to think it is.
      • Reply
      • Birdie commented: As someone who opted to donate this time around for the first time I'm definitely concerned. I support AO3 because they've been so good to writers, but allowing AI generated fic that literally learns from scraping fic written by real people who put their whole heart and soul into it isn't being good to writers by any means. And the bit about selling data to corporate for profit is *especially* concerning seeing how far above and beyond donors go during each drive to make sure the site stays up and running. This most recent one where they received more than $200k overnight was an excellent example of the dedication of donors. And the site is run by volunteers. Unless I'm misunderstanding something, volunteers don't get paid. So where does this AI-generated data profit go?
        • Reply
        • can't stop the wrock commented: certainly not to ao3! the site has already been scraped a million times over, and nobody on ao3s side will ever receive money for it. Just like reddit, twitter & co weren't compensated for having been scraped (which they have, also a million times over)
          • Reply
      • Nah commented: To be fair, being a chair doesn't really mean shit in the structural functionality of the org and AO3. It just means you carry out projects and make sure shit gets done. Now what she's doing outside the org, that's what sucks. Lawyers already meddle too much in artists' and creatives' problems, and the AO3's ToS are written around US law. Betsy Rosenblatt's support or nah of those laws is probably irrelevant in the grand scheme of things, but it still sucks.
        • Reply
      • Anon commented: Did they think we wouldn't notice?
        • Reply
  3. Plants commented: While I appreciate the clarification, knowing what the stance on AI from OTW as a whole would be nice and how you plan on protecting your userbase. AI does not belong in transformative works and I know myself and many other users are uncomfortable with AI being used to scrap our fics and to write fics to be later published on Ao3.
    4. Reply
    • Vilija commented: I have already reported a few "fics" that had in their summaries an explicit "written by AI for my prompts", and was told that they don't break AO3's rules...
      • Reply
      • omgtabby commented: I have also done reporting of these and been told they don't break TOS....
        • Reply
        • apprepuff commented: they dont break tos /yet./ if that petition gets high enough off the ground and OTW listens to us, we could be looking at a sitewide ban of AI-written “fics”, which would be the best thing OTW could do to protect individual creators, their works, and AO3’s userbase as a whole.
          • Reply
          • Nah commented: Are you talking about some change.org petition? Why would that change anything in how AO3 is run? You guys are funny if you think some "signatures" will change anything lol
            • Reply
          • Tina commented: And then what? People will not tag them as AI generated, so whose gonna prove they are? The detection mechanisms people use barely work, we have the same issue in academia. So you’re gonna end up banning folks who haven’t used AI based on gut feeling.
            • Reply
          • jillordinary commented: Next step: use complaints of AI generated fic to bully legitimate writers out of fandoms because a set of fans doesn’t like/agree with/disapproves of their works. I’ve read human created works that are indistinguishable from AI works, not because the AI is that good, but because the human work was not as polished or English wasn’t their native language. These fandom arguments are as old as fandom. Only the characters have changed.
            • Reply
        • g4egfqf3 commented: If AI fics were made against the TOS people would still make them but just not tag them, better to allow them so you can avoid the people who tag them.
          • Reply
  4. Taenith @ AO3 commented: Sincerely, thank you all for responding and for the clarification that Legal Chair Betsy Rosenblatt's atrocious views on the scraping (read: theft) of fanworks for AI text generators does not reflect the official position of the OTW. HOWEVER--- we need more than this brief statement that doesn't even mention the concept of AI. We need detailed action steps for how you are going to rebuild the trust you have broken with us, your donors/members as well as the fandom community you have been created to serve. MOST IMPORTANTLY, Legal Chair Betsy Rosenblatt needs to resign, effective immediately. *This is not over*, not even a little. Signed, a donating member of the OTW who has fanfic on AO3 and votes in Board Elections.
    5. Reply
    • apprepuff commented: i have nothing to add, this is just a good comment and i want people (especially OTW) to read it.
      • Reply
      • Taenith @ AO3 commented: Thank you apprepuff ♡ I really hope OTW leaders will read it too. It for sure *does* take time for any organization to respond to an unexpected outcry (I mean they absolutely should have expected it, but whatever. Regardless, something as explosive as Legal Chair Betsy Rosenblatt's conduct deserves a comprehensive and swift response, along with a heartfelt, detailed apology.
        • Reply
    • PhDelicious commented: Ditto everything in the comment above - including the donating and voting part.
      • Reply
      • Taenith @ AO3 commented: Thank you PhDelicious ♡ Our votes and donations keep the OTW functioning. I cannot believe we have arrived at this juncture, BUT--- let's see them try to pay the server costs if they keep going down this path. If nothing changes, the Fall Donation Drive will fail spectacularly >:(
        • Reply
    • Tina commented: scraping, as it currently stands, isn't theft. otherwise we wouldn't have cases like the stability ai lawsuit that rosenblatt commented on. her view, as a trademark lawyer, was this: > I would like to see courts consider the “training” process separately from the process of generating works. It is, of course, possible that a machine could generate an infringing work. But the process of training that machine involves something very different—turning expressive works into data and creating relationships based on that data collection. We call it machine “learning” for a reason. A well-trained machine won’t generate an infringing work, but it needs as large a pool of data to work from as possible to do that. The mere fact that an AI can create something infringing doesn’t determine whether the gathering of information is infringement. Consider the classic Sony v. Betamax case: The VCR can be used to infringe, but it has noninfringing (fair) uses, and therefore the VCR does not inherently infringe. I recognize that the analogy isn’t perfect, but I find it persuasive. In general, courts have found that “interim” copying isn’t infringement—that is, copying isn’t infringement when it occurs inside a machine and does not, itself, make copyrighted works perceptible to people—and I think courts should continue to follow that logic. That is, as the law currently stands, a perfectly valid opinion. Scrapping itself isn't infrigement, but it can generate infringing content. AO3 runs on the premise of "making derivative works off someone elses IP without consent". I do not quite understand why people want a volunteer of the organization whose entire job is to argue for a wide IP and Fair Use law to essentially shoot themself in the foot. You can disagree, of course, but scrapping away at Fair Use while not endangering fanfic will be quite the task.
      • Reply
      • WRD commented: "Scrapping itself isn't infrigement, but it can generate infringing content." This ↑ is an important distinction. Fanfiction archives might be the best - certainly the largest - source of how our culture currently structures and uses the written word. (As opposed to all the various work of say... 18th century Lit.) Teaching the algorithm how people communicate isn't the problem. It's the next step that may be infringing. It is the use of algorithm generated prose that is problematic, ethically questionable, and possible illegal. So then - are we at a place where we have to redefine exactly what IS a Transformative Work? Is a Transformative work intrinsically a piece of art that is both derivative AND generated by Humans using certain fundamental tools that could not produce a product of derivation on their own. (Such as, spell check, grammar check, style and clarity comparisons) Is publishing an AI generated work a form of plagiarism?
        • Reply
  5. Gina Marie Zuccaro commented: Ai definitely doesn't belong in transformative works, and should not be allowed to gain a foothold here.
    6. Reply
  6. Maud O'Bedlam commented: So what is the official position of OTW on this issue? Prof. Rosenblatt isn't just another volunteer, she's chair of your legal committee, so we have good reason to assume her opinions have sway in the organization.
    7. Reply
    • Nah commented: Why would you think that? "Chair" is a fancy word, but the responsibilities of them is to carry out projects and make sure shit gets done. The AO3's ToS are modeled according to US law, so even if she was anti-AI, it is the US lawmakers that have the final say. What sucks here is that someone so involved in fandom is pro-AI, but personally I think it's nuts y'all think she's somehow "powerful" enough to bypass everyone else in the org lol
      • Reply
      • Mazarin221b commented: Eh, she's there as a legal advisor, and as chair if the committee she does have sway over what issues are pushed, or not, and the framing of legal arguments from OTW. That's literally why she's there, to provide her legal expertise.
        • Reply
  7. Spell commented: Betsy Rosenblatt is more than just a member of a committee- she is the Chair of your legal division. That implies a level of power that I am not sure I am comfortable having an influence over AO3's future. I understand that this is a delicate situation and the Archive does not want to harm its relationship with their legal representatives but at the same time, those legal representatives need to understand the wants and needs of the archive. And that includes protecting the members from exploitation. to TLDR how this comes off from her original interview: So that would mean the people actually creating the work can't profit off of it, but major corporations? Go ahead!!!!
    8. Reply
  8. Yen commented: There should be no AI influence present what so ever because it’s theft. Please listen to us and do not implement it or you will see several people delete their hard work and leave your platform.
    9. Reply
  9. waddaluvr commented: I hope everyone who donated in the drive is able to vote for whoever could replace Ms. Rosenblatt. AI is unethical and I'm tired of people plagiarizing me and my friends fics via AI.
    10. Reply
  10. MJ commented: Interesting that you're minimizing the head of your legal committee as "one of [y]our 900+ volunteers." This isn't the act of one of your many many volunteers making ill-advised statements on their personal social media, it's your literal legal representative saying deeply concerning things on your official news output. Please take this seriously and hold everyone involved to account.
    11. Reply
    • miera commented: This is not correct. The chair of the legal committee is not the attorney for OTW or AO3, she is an advisor on policy matters. This entire thing is being blown out of proportion by people who do not understand the structure of OTW or what the different groups within it actually do.
      • Reply
      • well... commented: I believe it's fair to say that users no longer trust her advice, and that should matter
        • Reply
  11. Kekspeek commented: It is, in my opinion, pretty cowardly that the Board, the membership's elected officials who are the face and voice of the entire organization, are hiding behind this communications volunteer to deliver an "apology" while, at the same time, not acknowledging any of the very valid, very loud concerns both users and members have with AI and scraping the AO3. "We're working on it" is not a stance. Board of Directors: you've had time enough. Address the user base and membership directly.
    12. Reply
  12. TiroTiro commented: This isn't good enough. Tell is what your stance on AI scaping is, tell us how you will esnsure it doesn't happen, tell us why we, as writers and readers, should trust you now and give you or money?
    13. Reply
    • everything about this sucks commented: You can't ensure that scraping doesn't happen. They already slapped a DO NOT WEBCRAWL edit onto their robot.txt in January, and that's (sadly) literally all they can do. It's not effective at all, of course. Open AI and Co already pay poor people in third world countries like 2 Dollars and US based ppl minimum wage an hour to tag and sort through the material they have, and there's literally nothing stopping them from making every single one of those people an account and having them manually download all text on members-only forums etc. ChatGPT is already testing an always online mode (which will have access to any site open in the users browser), Microsoft, Apple & Co are all in on system-native AI as well, so there really is nothing that the OTW and or Ao3 can do that isn't entirely performative.
      • Reply
      • TiroTiro commented: Oh I know that. But they could still make it clear that they won't ever use the works on their site in AI, that they themselves won't be the ones who sell us out and that they're against Ao3 being used for AI. Obviously that wouldn't stop the theft, but at least it wouldn't be endorsed and encouraged by them and they could also make every work on the opted-out by default.
        • Reply
        • everything about this still sucks commented: > that they themselves won't be the ones who sell us out they ain't selling anything. not even userdata. they are a non-profit. Now am I for them throwing a giant banner on top of the site saying "LLMs cannot does not have emotions or creativity and can therefor not create fanworks -- therefor their work is not allowed on ao3" yes. Will that actually have an effect? No. It's gonna be as useless as our disclaimers were, will just make people post their LLM written crap untagged, and do nothing good in the end except invite more trolls to go around with their idiotic commenting bots.
          • Reply
  13. BiblicallyAccurateGremlin commented: Not a great statement considering how many people the article went through before publishing for editing purposes, and if this person does not represent the values of OTW than why the hell is she the legal chair???
    14. Reply
  14. Johniarty commented: This is lukewarm at best. AI has zero place on the archive. Period.
    15. Reply
  15. NO AI ON AO3!! commented: Nice non-apology! Betsy is not just any volunteer, she's your legal chair (a position OTW members can't vote on because she wasn't elected to her role). She clearly holds a lot of influence at OTW, otherwise you wouldn't feature her as you did! Please take an official stance on this issue: protect human writers! NO AI ON AO3!! NO SCRAPPING OF OUR WORK ALLOWED!! SET "OPT-OUT" TO DEFAULT!!
    16. Reply
  16. LW commented: I saw both the excerpt included by the OTW Signal and the ARL Views article that the Signal that it took the excerpt from, and I am more concerned about the contents of the excerpt. Why did the OTW feel the need to highlight the individual in question's views on the shippability of DALL·E and ChatGPT rather than the several paragraphs of content considering legal precedent of machine learning, and current debates about the definition of transformative work? That content might also be polarising... but it would have been significantly more relevant to a section summarising the individual's interview about AI legal issues. Definitely an odd editorial choice.
    17. Reply
    • Ring commented: Yeah. I actually would not have been on alert if they had published the relevant excerpt with an acknowledgment that it's a single person's views. But the first thing anyone into this stuff on an ideological level does is start using every vector they have to talk it up and any influence they have to spread support for it. If the org can't afford to lose her and she's deep into pro-LLM territory, it's going to become a real problem.
      • Reply
  17. RogueSareth says boot Betsy commented: This is not even close to a good enough statement. We want guaranteed work on better protections from Ai scraping for the archive(I understand it's impossible to completely stop it) and a ban on any obvious or tagged Ai generated fic. Also boot Betsy to the curb, I don't care if she's a volunteer, kick her out. A paragraph of "oh we're sorry you're concerned" isn't good enough when it comes to this even a little bit. You've completely shaken your users faith in this archive allowing someone with those views to be associated with the archive, especially when it comes to something as important as the legal side of things. Give us a real answer, and live up to the promise of saftey, protection and freedom for writers. It will be tragic, and a pain the ass to build another archive like this one, but we did it before and we'll do it again if it's clear it's necessary.
    18. Reply
    • RogueSareth can't believe you picked her to talk to the fucking government commented: Also, let's address the "one of 900+ volunteers" She is the chair of your legal team, no she is not just a random volunteer. And making that statement even more bullshit she was the official representative of OTW sent to speak to the US GOVERNMENT on Ai. That is not random and that does not lend credibility to her opinions not being a representation of OTWs official stance on Ai, which you HAVE NOT clarified
      • Reply
  18. One of the 900+ users who don't want AI commented: Then let's hear the official position.
    19. Reply
  19. Mekare commented: *Everything* about this statement is making me concerned especially in context of reading all the comments of people who were able to read the original news post. Leaves a bad aftertaste at best. Be clear and tell us what the organisation‘s official stance is!
    20. Reply
  20. FicVix commented: Wait, what happened?
    21. Reply
    • AshToSilver commented: https://www.arl.org/blog/applying-intellectual-property-law-to-ai-an-interview-with-betsy-rosenblatt/ - the legal chair did an interview saying she wants to have AIs trained on fanfiction.
      • Reply
      • MB commented: Thanks for the link. That's a very lawyerly opinion, most artists are against AIs and becoming chopped liver.
        • Reply
      • Nah commented: Betsy Rosenblatt never said she "wants" AI to be trained on fanfiction, she's aware that it's already happening and she thinks it's cool. Let's critize her for what she said, and not put words in her mouth, hm?
        • Reply
      • Tina commented: She doesn't? She says that she seems some benefit to it (AI models picking up more modern ideas compared to when they were mostly trained on older works) and that legally she thinks there needs to be a distinction between the training vs. generating part of AI (i.e. scrapping and training isn't inherently copyright infrigement as it currently stands, but it can be used to create non-fair-use content).
        • Reply
  21. Dreamin commented: We need to know AO3's actual stance on AI. Also, it wasn't just one of the 900+ volunteers, it was the head of the legal department. This vague response is not good enough.
    22. Reply
  22. Longtime AO3 user commented: I appreciate the clarification. However, she is not quite "just one of 900+ volunteers". Ms Rosenblatt is the legal chair of the organization, or at least was identified as such. That leads us to believe she is someone who wields some level of power in the organization, which is what made her comments all the more concerning. We would appreciate further clarification and an official stance on AI scraping and generating of fanfic some time in the near future.
    23. Reply
  23. RaraeAves commented: This is cotton candy fluff of a non-statement. The fact that the head of the AO3 legal committee, which is not an elected position, seems so cavalier about writers having their work stolen via scraping - during the WGA Strike, no less - has me sincerely re-evaluating my continued support as a member, a donor, and a user. This is unacceptable. What measures is AO3 willing to take for people to feel secure in their work staying on the site, if any?
    24. Reply
  24. Kamari commented: So you're definitely getting rid of Betsy R., your Chairmen of the board of Legal, who voiced open enthusiasm for AI scraping of Fanfic (which is theft), right? Youre going to make clear your official stance on AI right?
    25. Reply
  25. AshToSilver commented: Seconding everyone x1000 - we need additional clarification, this is a serious issue and Betsy isn't just 'one of 900+ volunteers'; she's the chair of the legal team and can't even be elected! Sweeping everyone's concerns under the rug is NOT the appropriate action here.
    26. Reply
  26. Marithlizard commented: Thank you for the clarification. While I am concerned as well - and I think we're all very clear that no one wants their AO3 works to be used in training AI that corporations use to avoid paying writers - I understand that you can't instantly produce an official policy statement out of your hats. The outraged calls for blood here seem a little premature.
    27. Reply
    • k commented: Except the article was published a week ago, not today. Not to mention that people have been asking AO3 to clarify their stance on AI for months now, as well as reporting AI written fics (both to no results)...
      • Reply
    • Plantzawa commented: Significant points of concern are that these views came from OTW's Legal Chair, a position that is NOT elected, and that this interview was actually conducted in February, well before the annual donation drive, but they only chose to release it AFTER receiving everyone's money. Legal enthusiasm for widescale theft from an organization that historically has been a bastion of protection against such action, while using that theft to feed machines that make commercial profits, something fanworks cannot do, is DEEPLY unethical. Releasing the interview after AO3 made their yearly donation income indicates that they are fully aware of what they are doing.
      • Reply
  27. Elena commented: Hi guys! I'd like to know what you are doing to ensure you provide users with appropriate information about an issue that affects so many of us, rather than just assuring us you're "taking it seriously". I'd also like to know if you're reevaluating how committee chairs conduct themselves in public spaces, given that the interview presents Rosenblatt as an authority within the organization, and indeed that as a committee chair she has a formal role and is considered a staffer. I think we all know this isn't the same thing as a tag wrangler posting about AI. It's disrespectful not to acknowledge that! And finally, I'd like to know how you plan to reach out to users regarding AI given that Rosenblatt has formally stated, in her capacity as a representative of the org, before the Library of Congress, that she's "heard from" users regarding AI.
    28. Reply
    • Ring commented: I am concerned especially that this may be someone whose volunteer work they depend on and can't replace. I would like to see a statement from both the OTW and Rosenblatt clarifying that her position is not being used to capture an audience for her personal opinions on AI. If they can't provide that, she has too much influence as "one of 900+ volunteers."
      • Reply
  28. DL commented: This feels a bit like throwing Betsy under the bus. On April 19th, Betsy spoke to the US government on this topic as a representative of the OTW. https://www.copyright.gov/ai/listening-sessions.html Her statements in that listening session seem congruent with her statements in the interview that OTW Signal linked. If Betsy's professional or personal opinion on the subject conflicts with tbe OTW's official position, why is she being sent as the OTW's representative? Putting her forward as an official representative, but then saying that it's just her opinion when fandom reacts negatively, seems unfair. What IS the OTW's actual position, if it's not what was presented to the government? How has that position been arrived at? Hiw was the membership polled?
    29. Reply
    • Sparky commented: THIS is what I’m worried about and what I think has been under acknowledged in the wider conversation on Rosenblat’s statements. I find it extremely concerning that she is expressing similar views while actively functioning in her role as the OTW’s legal representation. If she is “just” one of 900+ volunteers and her views are not representative of the organization’s stance, why has she been allowed to air said views in an official capacity?
      • Reply
  29. ripeteeth commented: Removing the excerpt does not change the fact that the legal chair of the OTW is sympathetic to using ao3 AI training, a stance I (and many others) are extremely uncomfortable with. While this may not be an official poliy, the removal of the statement and this update do nothing to settle my concern that the OTW is interested in pursuing this further or to repair my trust that the current leadership is putting the best interests of the userbase first.
    30. Reply
  30. Ring commented: I appreciate that this is an incredibly difficult situation and that internal conversations are likely ongoing. Please understand that this one person's opinion would probably not have prompted the overwhelming response it did if people hadn't been waiting for serious communication on this topic, only to have this come out as though the conversation wasn't happening. I don't actually think you all will do anything to integrate AI into the platform; I do find it concerning as hell when a nonprofit has someone in what is probably a very necessary, influential, difficult-to-retain position casually pushing LLM shit to an audience that has made it clear they want nothing to do with it, and the org publishes it uncritically with a cutesy comment. It's concerning specifically because people who are crypto, Web3, or AI true believers DO NOT STOP PUSHING THEM, NEVER SHUT UP ABOUT THEM, AND WILL USE ANY PLATFORM THEY HAVE TO TRY TO LAUNDER PUBLIC OPINION OF THEM. I want to know that there are not people on the board with LLM-based investment portfolios who are going to get to turn their entire fucking focus here into "what if fanfic can stop the bwain-having wobots from being bigots???" and use official comms to write their own fanfic about it, because that's what every single person does when they're more in touch with a tech-utopian fantasy than actual people.
    31. Reply
    • RogueSareth commented: This right fucking here
      • Reply
    • Victoria commented: exactly! exactly. this comment gets it.
      • Reply
    • This commented: Banger.
      • Reply
  31. AO3 LoveIsStrong commented: Thank you for listening to us. Please provide your official position and please ensure fanfics are never used to "train AI" without explicit author consent. As of now authors of 69 works across 22 fandoms have joined hands. OTW must clarify their official stance. Our labor of love cannot be harvested without our permission for corporate profits. We support WGA Strike. We are against unregulated AI. https://archiveofourown.org/collections/NO_AI_Training
    32. Reply
  32. apprepuff commented: i said it on twitter, and i’ll say it on here too: ❝ something i think everyone needs to remember is that AO3 was never meant to only be for fanfics. the site was always planned to be an archive for all kinds of fanwork, including fanart. if we let AI “fics” slide, then AI “art” isnt far behind, and we all hate that shit just as much. WE NEED TO SHUT THIS BULLSHIT DOWN WHILE WE STILL CAN. ❞
    33. Reply
    • let's just go back to paper zines lol commented: WE NEED TO STOP THIS WHILE WE STILL CAN -> that ship sailed in 2019.
      • Reply
      • hey paper zines are cool commented: Says you! People said the same thing about the ~metaverse~
        • Reply
  33. 2501 commented: except she's not just one of your 900+ volunteers, she's the head of legal and is evidently pro-AI. This is a weak, weak statement and it will affect more than just your donations next year, so you best start coming up with a more derisive statement on why this timing, and why these statements from this particular legal rep.
    34. Reply
  34. e commented: Everybody who donated before June 30 can vote in election August 11-14. Make your voice heard and kick out Betsy Rosenblatt. Apologists of theft don't belong at OTW. As for OTW "statement" - embarrassing and cowardly. You should clarify your stance on AI immediately. If you really want to defend writers - then you should ban AI. If you want to defend poorly disguised corporate interests and theft - then prepare for AO3 to perish in a year, because no one will donate to a site that encourages plagiarism and no one will post new fanfics there. The real statement should be ban of AI and any "works" regurgitated by it.
    35. Reply
    • Nah commented: Betsy Rosenblatt cannot be "kicked out", she's not an elected official of the OTW Board of Directors. She's just a chair. Yes, the legal chair, but just a chair at the end of the day.
      • Reply
  35. onereyofstarlight commented: Having now read the original interview referenced in this statement, I too am concerned. While it is true that machine learning does and should involve a wide variety of sources, one of the key differences between machine learning AIs in recent months is how the data and information being "fed" to the programs is being gathered. Machine learning used to be in an extremely controlled environment, using data carefully selected by the programming and engineering teams involved, now it comes from scraping the internet with no chance for permission or acknowledgement. This is the key issue - unlike our transformative works where we fully acknowledge the influence and works of other authors, no such ability exists with AI. Furthermore, the method of data scraping when many of us have experienced violations due to privacy in other spaces of virtual data in other areas is genuinely concerning. I would add my voice to the call of an official statement regarding OTW and subsequently AO3's position on AI in two feilds: 1. Your position on AI scraping fanfic from AO3 specifically, and 2. Your position on hosting AI generated fic on AO3. We need to be able to make an informed decision regarding how our work is hosted as this is a big decision for many of us about whether or not we continue to use AO3.
    36. Reply
  36. longtime AO3 advocate commented: The excerpt wasn't the problem - it's Rosenblatt's position on AI as your legal chair that's the problem. Also, this "update" is really, REALLY insufficient: you haven't mentioned at all that the issue is about AI, and you also haven't given your official position. You need to. In fact, you need to acknowledge the topic of the excerpt as well as Rosenblatt's position, and refute it properly, otherwise your look is ... not great. Anyway, just to keep things above board, here's the link to the interview in question (which is the actual issue), in case people missed it and no longer have the link you deleted: https://www.arl.org/blog/applying-intellectual-property-law-to-ai-an-interview-with-betsy-rosenblatt/
    37. Reply
  37. veryroundbird @ ao3 commented: Thank you for the timely clarification. While I'm not necessarily invested in Ms. Rosenblatt's resignation, I would like to hear from her that she understands the concerns of the userbase about generative modeling as the legal chair of the OTW. While, yes, it is her own non-controlling opinion, and she is one of many volunteers, she will almost certainly have to be working on measures around the legality of fair use and generative modeling. I hope a stronger and more detailed statement is in the works, and I'll look forward to it.
    38. Reply
    • DL commented: These sentiments have already been conveyed to the US government regarding fair use and generative content https://www.copyright.gov/ai/listening-sessions.html Betsy was acting as the official representative of the OTW in that hearing, which makes it seem unfair that the OTW is disavowing her interview as her personal opinion now.
      • Reply
  38. SN (AO3 user and donor) commented: This statement is insufficient given the number of concerns raised by the members of this community. Betsy Rosenblatt isn't just any volunteer, she's not even just any legal committee member, she's the CHAIR of OTW'S legal committee. She was the one who represented the entire organization before the US Government Copyright Office. The opinion she gave in that interview was irresponsible. And her statements at the hearing on April 19th are not entirely clear about the organization's stance on AI. Her closing statement mentions: "We've heard of enthusiasm about the potential for what AI can do and bring to fan communities." I know that AI in the context of tools like text to speech/alt text for the visually impaired and speech-to-text for the physically impaired can be beneficial to the community, for example. And that way they are welcome. But here we are not even talking about "transformative works". It's theft and regurgitation of styles/content, without consent. Indiscriminate web scraping harms our community. OTW needs to be clearer on its position, especially now that AI regularization and limitation is being discussed. For now I will not upload more new works to the site, until we have a concrete statement of the organization's position, my future donations will also depend on this. And, in future elections, I'll pay attention to the position that each candidate takes regarding AI.
    39. Reply
  39. fanficreader105 commented: I'm sorry but this hysteria - no other word will do - is getting out of hand. Asking AO3 to ban AI while giving no indication of how such a ban could be implemented is just absurd and performative. As for threatening to leave AO3... Do us readers a favor and leave already instead of clogging comments - the world of AO3 is large and you will not be missed.
    40. Reply
    • apprepuffl commented: tell me you don’t understand the gravity of the situation without telling me you don’t understand the gravity of the situation
      • Reply
    • DL commented: I don't see a way for AO3 to "ban AI." The OTW can't stop scrapers, can't remove content from training datasets, and can't detect whether or not a fic was written by AI if the person posting it doesn't confess. But that doesn't mean the OTW has to say that scraping is good for fandom or encourage the use of fanfic in training LLMs.
      • Reply
    • TiroTiro commented: We know it can't be banned because then it will just go untagged, that's not what we're asking. We're asking them to clarify their stance, to promise us that they won't use our work and sell it to others to train data with and monetize something which they themselves are very clear they don't allow to be. If writers can't put their ko-fi or tip jars because its non-profit then the Archive shouldn't encourage mass scraping of the Archive in order to profit from these works. Also yes, there's 11 million works on the Archive, hard work and passion of tens of thousands of writers, a lot, however what this has made clear is that a good chunk would rather delete their work then be used as AI data. It might not happen all at once but it will be a domino effect and pretty soon all Ao3 will have is AI generated works.
      • Reply
  40. Piperderg commented: Removing the excerpt does nothing and instead makes it look like you are hiding the issue. Removing Betsy, on the other hand, would be a better solution. There's a difference in AI being used to correct grammar (or spelling) or speech-to-text or text-to-speech for disabled writers. those types of AI are not the issue.
    41. Reply
  41. Very Concerned commented: While I am grateful for the clarification, classifying your literal legal chair as just one of 900 members feels a little disingenuous to me. I am also concerned that you do not, in this message, take a unified stance against the use of AI on this platform. This does not really abate my nerves about the situation.
    42. Reply
  42. MangoTea commented: I am glad you made a statement but this is not sufficient since Betsy Rosenblatt is your legal chair. She is not simply one of the 900 volunteers. I know that the scrape has long since happened and in and of itself it's likely legal, but that embrace of corporations monetizing our work in sketchy ways was chilling. Also, output from these machines is imitative, not generative and basically slightly obfuscated plagiarism. Also, these imitative machines don't run themselves. Commercial applications involve exploiting labor in the global south, which is one of the reasons the WGA is against them. John Rogers, whose tweet is linked below, is on the WGA AI policy work group. Please familiarize yourselves with the content of this thread and also the WGA's overall policy and reasoning for it. This committee and Betsy in particular really needs to educate themselves better on what's going on with these imitative systems using the term 'AI' as a marketing buzzword https://twitter.com/jonrog1/status/1653415215915229184
    43. Reply
  43. AU (long-time member & former donor) commented: I hope everyone in the communications team understands specifically how frustrating it is that the head of the Legal team has time to do a fluffy, pointless interview with ARL but can't be bothered to communicate directly with members about critical issues that have been raised for a long time. It is also frustrating that she is representing the OTW to US Congress while having such a poor understanding of how LLMs work. There is nuance to the "AI" issue and this could be used as a moment to educate OTW users on what has and can be done, but it seems more likely that you will simply alienate a significant number of people who otherwise have been staunch supporters of the organization. If you think the anti-AI segment is merely a loud minority and you're fine with unauthorized scraping of the site then make your argument clearly and directly and let people decide. This issue has already spiralled into hysteria on social media in the absence of any real comment from people with authority within the org - and it is not the first time this has happened. Get your shit together.
    44. Reply
  44. amarisllis commented: This is a start, but what we are really asking for is clarification from OTW on your organizational stance on AI, and I for one (as well as most of the community) would like to see the organization take a firm negative stance against allowing AI training on fan works and allowing AI generated works on the platform. Anything less and I will be seriously re-evaluating future support for the organization both financial and otherwise.
    45. Reply
    • amarisllis commented: Wanted to clarify my earlier comment here. After thinking on this further, while i would like to see rosenblatt step down as legal chair and the board take a firmly legative stance on training ai via fanworks, im not sure im comfortable with a ban on ai generated works on the platform UNLESS the board has a firm plan in place to ensure that human fics won’t get caught in the crossfire. Im not convinced rhat ai detectors are good enough yet to ensure perfect classification of ai text and given that caveat i would be hesitant to see a full ban. However, i would still very much like to see clarification on what ao3 has already done to stop scraping (since im aware that they have already modified the site code to block common scrapers).
      • Reply
  45. v commented: Your legal chair spoke to the US copyright office, representing *your* position and calling AI output transformative and exciting, when it's actually just plagiarism and mined from unconsenting writers many of whom had hoped you'd represent them. This is so galling. We need official, clear OTW's stance on AI. We've been waiting for months for you to say something about the illegal scraping of your website to feed these AI models and how you're going to represent these valid concerns to authorities and THIS is your response? Disgusting. You need to tell us whether you're actually going to represent an anti-AI stance to authorities going forward. Because if not, our future support stops here.
    46. Reply
    • v commented: Questions the OTW needs to officially answer- 1. Does the OTW consider AI output to be transformative? (Hint: it's not since it's a generative algorithm) 2. Does the OTW consider it fair use? (Hint: it's not since it directly monetarily competes with the artists whose very work it ingests) 3. Will the OTW campaign for the complete removal of AO3 work from commercial datasets since it's in direct contradiction with the fact that you believe fan work that makes *no* money is fair use? Please, answer these questions.
      • Reply
  46. Concerned commented: While i appreciate the clairification, Ms. Rosenblatt isn't just a volunteer-- she is the Legal Chair. Her outlook does have sway within the organization, and I would like to see a statement from OTW that they will do whatever is within their power to disallow AI to be trained on works within the Archive. I hope that this is just the first of many statements to come. However, if this is the only statement planned, this is not enough.
    47. Reply
  47. moljn commented: I don't have an opinion about the interview itself, but she's your long-term legal chair, she's credited as such in the article, and she's talking about fans, fan works, and transformative works. She's not a random volunteer, and trying to sell her to us as such is insulting. I mean, come on. If she's not speaking in her OTW role, she (and you) should have made that extremely clear in the interview (and the linking to it).
    48. Reply
  48. spockina commented: "That article featured the opinion of one of our 900+ volunteers" is not how I expected you all to handle this. That article featured the opinion of your Legal Chair; that's not just a volunteer out of 900, and this approach is extremely disrespectful. Don't try to downplay what happened. I really hope you guys work on something that is more honest and open, and that actually looks into the best interest of your users.
    49. Reply
  49. kelex commented: Given that AI scraping/training is already happening, the only way for AO3 to actively stop it is to... shut down. Delete all our works, leaving nothing for the AI to scrape. I don't know that any of us want that. As long as anything is online, it is accessible for scraping. There are probably stopgap measures (making fics available to logged-in members only, is one I've seen kicked around, I don't personally know how effective it will/could be.) that might be temporarily successful, I can't imagine how there's going to be any significant changes that won't completely impact the way that AO3 operates. Banning AI-written fics or requiring disclaimers is only going to discourage those disclaimers from being used, and allow AI-written works to slip easier into the mix. AI needs to be banned from creative endeavors, and AO3/OTW does need to take a hard stance of "We do not agree with this at all, and we will try and discover how to stop it" but as a user base, WE have to realize that there is, in all likelihood, very little that we can actually do to stop it unless we go back to mailing things to each other in paper zines and keep it completely off the internet forever.
    50. Reply
  50. solrise commented: Others have said it before me, but for what it's worth to add my voice to the pile: Claiming that Miss Rosenblatt's stance does not reflect the OTW's is patently false when the current response to reports on AI generated fic on the archive is to inexplicably claim that the work is not infringement unless it contains, in part or full, word-for-word text from another source, on or off the archive. This definition of plagiarism is too narrow and simple minded in the current technological climate, and the archive's ToS is in desperate need of updating to reflect the sane definition of plagiarism as "passing another person's ideas, thoughts, or words off as your own without accreditation," which generative AI demonstrably violates. If a college student can't get away with it, neither should generative spam on an archive that touts itself as being built by and for fans. Secondly, trying to claim miss Rosenblatt is but one one many volunteers is misleading and disingenuous. Her words hold far more weight than the average OTW volunteer, as her position is not an elected one and she represented the OTW before the United States Copyright office. There is serious and lasting consequences to her position that need to be properly and fully addressed, not swept under the rug and disregarded. I join the others in saying that her position has demonstrated a clear conflict of interest, when she sides with Stability AI and other generative AI technology companies against creatives who have had their intellectual property rights grossly violated. While the fast response is appreciated, it does nothing to actually quell the valid concerns I and many others have. Do better.
    51. Reply
    • Tina commented: College students are getting away with it, because AI detection does not work (believe me, I run their essays through it all the time. I've run my own stuff through it, and it showed up as AI generated). If the TOS are updated to include AI-generated work all that would do is stop people from tagging it and leaving moderators to essentially guess whether a work is AI generated or not. How does that, in any way, help the situation?
      • Reply
    • Tina commented: Also, banning "passing someone elses ideas as your own without credit" on a fanfiction website??? you want me to credit that user on the j2 kinkmeme every time i write abo?
      • Reply
  51. S commented: She's not "one of our 900 volunteers," she's the LEGAL CHAIR. You letting her run her mouth about AI is proof you agree with her, and you're only now throwing her under the bus and saying it's "just her opinion" because we're mad at you. You people are utterly irresponsible. I had already lost a lot of faith in you because of your slow action (basically inaction, let's be real) regarding racism on the archive, and this is just the cherry on top. You people don't give a SHIT about your userbase. I don't even know what you stand for anymore.
    52. Reply
  52. l commented: ok and........... what are you doing abt it? you can't just put up here and have her do an official interview somewhere else being neutral/for ai. dont just say "ummm well its 1 person (WHO IS ON THE LEGAL TEAM) sorry guys", do better.
    53. Reply
  53. blueskyscribe commented: "The opinion of one of our 900 volunteers" is a really disingenuous description. It makes it sounds like this person is a part-time tag-wrangler. In actuality she is the chair of OTW's legal committee. "It's not the official position of our organization" is what Subway says when one of their sandwich-makers posts about how aliens are living inside the hollow Earth. But it's a strange response after OTW voluntarily boosted portions of this person's AI interview. If OTW supports AI then I will no longer be donating to it.
    54. Reply
  54. sil commented: I've seen countless companies and organizations try to sweep concerns raised by people who use their services under the rug because they need these people nice and silent so that they can go through with their unwelcome plans without any disruptions. This "explanation" here feels a lot like you're trying to do exactly that.
    55. Reply
  55. AvidBeader commented: I would rather have seen you take a few more days to consider your message before putting this out. It is blatantly misleading corporate-speak to refer to the chair of your Legal Committee as "one of 900+ volunteers", instead of a position of leadership within the OTW organizational structure. I can appreciate that the OTW board may not have a policy finalized and ready to share regarding the use of AI in hosted works or the questionable ethics of allowing fanworks by those who cannot legally profit off them to be scraped by businesses that will profit. I can also appreciate that there is only so much that can be done while we wait for US law to catch up with tech innovation. But please consider: (1) ensuring to the best of your ability that fanworks cannot be scraped going forward - stopping Common Crawl was a welcome move, but I'm sure more are around or will turn up. (2) Make "AI generated" an Archive Warning along with the Big Five, and wrangle the various tags so users can filter it out simply by excluding that phrase. I don't expect it can be kept out right now - trying to ban it completely will result in some people not tagging it. It's going to take the same kind of pressure that artists have brought to show that "writing" with AI is not acceptable in works of fiction (or frankly anywhere outside of service industries, but we need to focus). Finally, I do think Ms. Rosenblatt needs to step down as chair of Legal - her views clearly do not align with the vast majority of your user base and we expect someone in her position to be looking to continue protecting transformative works instead of opening them up to misuse by tech toys.
    56. Reply
  56. Daeg commented: What's the official position? It better align itself with what the archive's creatives agree with. Or the archive will be archiving cobwebs. We won't stand to have our works used as fodder for a machine. This isn't animal feed, it's countless hours of hard work, freely shared with other creatives and the rest of the fandom. If people want to feed AI, then create their own feed for it. Don't steal other people's creative work. AO3 is an ARCHIVE, not a feeding ground. It houses stories, texts, books, not "nourishment" for a machine that is being used to kick writers out of their jobs with poorly fabricated, plagiarised texts.
    57. Reply
  57. green commented: I have been active on this site--reading, posting, donating, and voting--for *over a decade*, and I absolutely will remove all of my writing and every comment I can find before I'll let you scrape them to "train an AI." if I get a whiff that such a scheme is in progress I will act immediately. I am beyond furious about this. bad enough that you're apparently approving of AI-generated "content" already; I hadn't known about this, and now that I know I am strongly reconsidering any future donations.
    58. Reply
  58. MinMu commented: I am very concerned about this issue. At the last election I was concerned about candidates who were promoting a change in the direction of AO3 and /most/ of those problematic candidates were not elected (but it required a lot of people to explain how to vote and why those candidates wanted to move in another direction.) I *like* the foundation of this organization and I support it financially. It is offensive to minimize and that it is not important that the legal chair wants to change the direction of the organization. I know that you (wonderful, hardworking!) volunteers are fighting other issues on other fronts, but don't let this issue be ignored because you don't have the margin to tackle it.
    59. Reply
  59. Dez commented: What a pathetic apology. The best part is that no where in it did you promise to not let AI scrape the archive, despite the reaction you guys got. That's my absolute favorite part. That you actually have the audacity to spout nothing but pretty words to WRITERS. Incredible. So until you guarantee us that you're not going to let AI scrape the archive, while also backing up this claim with irrefutable proof, I will forever know that you are going to feed our writing to the robots, with or without telling us, because you are showing all the signs of corruption. I also firmly believe that you guys have already saved back-ups of every story on your servers (text takes up very little space; 64,389 characters in a Libre Office text file is only 60 kilobytes), or at least every most-popular one, in preparation for people deleting their works in retaliation for this GENIUS move you've made.
    60. Reply
  60. KineticKid commented: Plagiarism is plagiarism, no matter if it's another person or AI bot of some sort. Should not be tolerated either way. Too many of the writers on this site have put their heart and soul into making these stories for anyone/thing to be allowed to basically spit on it. That's wrong. Any way you look at it.
    61. Reply
  61. Former donator commented: Is this the "one of [y]our 900+ volunteers"? The insignificant, not-noteworthy volunteer who has headed your legal committee for a decade and represented the OTW at the US Copyright Office session on AI? It sure sounds like she's representing the OTW's opinion on AI. "In April, Legal Chair Betsy Rosenblatt participated in the first of the U.S. Copyright Office listening sessions on artificial intelligence and copyright law. This listening session was a great opportunity to provide information, help frame U.S. lawmakers’ questions about future policy decisions, and learn about technology and practice so that the OTW can make its own policy decisions about how to approach questions about AI. Legal will be doing more in this area in the coming months." "https://www.transformativeworks.org/april-2023-newsletter-volume-177/ I will no longer be donating to AO3 in the future. The handling of this matter is absolutely deplorable and does not give me hope for the future of the Archive, Fanlore or the OTW as an organization.
    62. Reply
  62. K commented: While I appreciate the 'brief' explanation, I think you are underplaying the position that "volunteer" holds with the A03, which is disappointing. We need a very clear cut explanation on (1) if AI will be implemented on A03, (2) if AI created 'writing' will be allowed, especially if it's created with a program based on the basis of other writers hard work and creativity, (3) how you will protect our work from AI programs and thievery, (4) when rules will be put into place that clearly state if you will or will not allow AI in any way. We need this so we can make an informed decision on: If we wish to keep our work on your site. If we wish to continue to support a site that is funded by us to protect our work - but might potentially be breaking that trust by allowing AI thievery. What steps we need to take to protect our work, if you fail to do so by allowing AI onto the site, especially if it's for PROFIT. Because, remember what is written on your site: "The OTW is a fan-run nonprofit dedicated to the preservation of fanworks. We're sustained entirely by donations from our users!" Allowing AI to steal from writers is not protection or preservation. Why would we want to fund that with our donations? Thank you.
    63. Reply
  63. M commented: AI is currently unethical, in my opinion. AI like ChatGPT can be used to devalue the time and effort humans put into their creations - their writing, their art. I would greatly appreciate knowing Ao3's stance on AI taking writer's stories to reproduce similar (if not the same) ones. Even if an individual's opinion is different from the Ao3 stance, I want to know whether human effort will be valued or not by OTW/Ao3 in official/practical terms.
    64. Reply
  64. Ruro commented: What we want as creatives who use AO3 is firm confirmation that it will not ever be using AI, and take a stance against it. This is pathetically mealy-mouthed. As someone who's donated plenty in the past and is a voting member, I sure would love to be able to vote on whether or not AI should get to steal my creative efforts.
    65. Reply
  65. Scorpionturtle commented: Hi, Can you clarify a few things- you mention this is "just" the personal opinion of one of the 900+ volunteers but you choose to publish that personal opinion on your blog so how is that not an endorsement of her views by the org or at the very least promoting them? Also she is identified in the article as not "one of 900+ volunteers but as " legal chair for the Organization for Transformative Works (OTW), a nonprofit dedicated to preserving fan works, which form looking at your structure is not a position that is voted in but is one that has a significant role without any mechanism for the voting members to remove her. Clearly since you choose to publish the interview you knew she was identifying as a representative of the organization so again how is this just a personal opinion. Why remove the excerpt when the person who's views the excerpt was when she is still in the same position in your organization? Her views and roles have not changed and the interview still exists...the only reason for removing it is if you are trying to cover this up which isn't a great look and is rarely an effective means of actually dealing with an issue of a core volunteer who has influence in your org having beliefs that are directly the opposite of the vast majority of your voting members. This response is disingenuous at best and pure hogwash at worst. Please respect your voting members and the larger community to not promote AI as a positive impact on fanfic on your blog or in policy and not have volunteers who endorse that speaking as an identified representative of OTW.
    66. Reply
  66. Lymmea commented: As a longtime AO3 user and donor who's only just heard about this, I've been doing a bit of reading of the comments to orient myself. And while it seems like Rosenblatt's being the legal chair is not necessarily that significant in terms of how much power she has over deciding policy, I do think that identifying her as only "a member of the legal committee", referring to her as "one of our 900+ volunteers", and not clarifying to people outside the organization (like myself) who aren't clear just what being the legal chair means or how much sway she has in the OTW was a pretty big misstep. The comments are alarming, and rather than downplay Rosenblatt's standing by not even mentioning it, it would be better to clearly define her role and - if she truly doesn't have influence over the OTW's overall AI policy - explain that directly. Unfortunately, the wording of this announcement feels disingenuous by making it sound like Rosenblatt is just a face in the crowd of the organization when her position at least makes it SOUND like she's a lot more than that. And yes, it's very possible people are blowing up her title to mean more than it really does. But I do think the OTW left itself open for misunderstandings like that by not explaining her position to its users, many of whom it should recognize are not familiar with its organizational structure and titles the way people on the inside would be. I feel like this is a failure of looking at things from the perspective of the average OTW user and communicating clearly with us. While I recognize there is very little the OTW - or anyone, really - can do to prevent scraping of people's work on the internet by AI, I do agree that such practices, and encouraging the sorts of tools that make use of real people's creative works in order to make machine-generated amalgams of creativity, is pretty abhorrent. I wouldn't be against clearly labeled AI-generated works on sites if there were currently any ethical AI programs (that I know of, but as far as I know this is the model they're all based on) that DON'T rely on being trained by, and profiting off of, real people's artistic endeavors. But, to my knowledge, that sort of thing simply does not exist yet. If there are any AI programs exclusively trained only on royalty-free, "anyone can use this work for any reason" creations/on their creators' own works, or if there are any AI programs that are taught in different ways, I've never heard of them. So while there may be interesting possibilities for different kinds of AI programs in the future, the ones we currently have that generate creative works are all unethical in nature. If the OTW's policy is against plagiarism committed by humans, which I assume it is and certainly ought to be, then I don't see how there can be a question of what its policy should be on AI works. AI-generated works, when those programs are based on material stolen without consent from human creators, is just a plagiarism hash. No matter how finely the material is chopped up so you can't recognize the origins anymore, we nonetheless know that the origins can be traced back to works the AI did not seek or acquire permission to use for its or its creators' benefits. Content creators of just about every sort and stripe, even those of us who can't and don't seek to profit on our works ourselves, generally have no interest in our works being used without our consent to help some programmer profit. In fact, most of us feel outraged and disgusted by the idea. Perhaps as a legal advisor, who may not be a content creator herself, Rosenblatt has a different perspective on these things...but then perhaps her perspective shouldn't be one you amplify, in that case. Indeed, choosing to amplify an opinion - even just an individual's opinion that does not dictate or reflect the OTW's policy - that is pro-AI scraping when the subject is so contentious, and the OTW has no defined policy on AI one way or the other, feels like if not taking a stance on the subject purely based on what kinds of views the OTW is putting out there, then at least testing the waters. (If Rosenblatt was bait to see how people would take her comments, I feel sorry for her being made a sacrificial lamb.) Whether or not this was intentional, you can't deny how bad it ended up looking. The OTW has always followed US law in terms of what it does and doesn't allow, and I understand that. With no laws against AI works that I know of on the books yet, I can understand if that's the way the OTW leans on the matter, although I may not like it. But I also don't think making certain provisions to prevent issues that specifically affect archives like AO3 would be entirely out of line. US law is unlikely to legislate ever, I think, on plagiarism of nonprofit fanworks, machine-assisted or otherwise. Plagiarism in education, for instance, is generally dealt with by the schools, not by the government. And again, as far as I know, plagiarism is already not permitted on AO3 when it occurs between humans. By all those metrics, the OTW taking a stance against AI-generated works, for as long as they are created using stolen content, seems like a self-evidently necessary measure. (I don't think the OTW would necessarily be out of line to dictate that its works all be created by humans, regardless of whether or not the scraping of stolen works is involved, but I think that may be more of a subjective issue where there could be interesting arguments made on both sides.) But at the end of the day, your supporters are simply asking that you make it clear that you do not support, and will not host, works that are plagiarized. Whether one work was stolen wholesale by a person, or a thousand works were dismembered and reassembled piecemeal by a computer program, should not make a difference in the response. I acknowledge that there are probably going to be significant difficulties in enforcing an AI content ban - how to recognize and remove AI-generated content, if it's even possible to prove a work is AI-generated definitively, how to avoid false-flagging poor writing or language barriers as the result of AI confusion, false claims of AI use from users acting in bad faith, etc - but I suspect that for many, it's less about preventing AI use on the site itself (which I think those of us who are reasonable and understand the logistics understand can't be easily done) and more about the OTW having a clear ethical stance in regards to whether it welcomes or condemns the use of programs that steal from us for profit we'll never see.
    67. Reply
  67. Lymmea commented: ...apologies for the massive paragraph above, I had no idea my paragraph breaks were not going to translate.
    68. Reply
  68. Carlos_Of_Night_Vale (you know, a writer. on ao3) commented: the wga is literally on strike because big media companies are refusing to regulate ai and pay them fair wages. so after a week of striking, your legal chair wants to do the exact thing to fanworks that professional writers are striking over and you also refuse to allow writers on the site to ask for tips for writing literal novels when you make thousands of dollars multiple times a year? be so fucking for real right now
    69. Reply
  69. R commented: This post tells us and clarifies nothing. It's concerning that you could have avoided the majority of this upset--and most of the misconceptions causing it--if you'd dedicated yourself to communicating clearly and honestly the moment the issues with AI-generated content cropped up. You need to take this as what it is, you breaking our trust in your promise to always side with fanfiction writers, and start working on gaining it back. I understand that you can't keep works safe from crawling and already implemented what could be done, but your negligence to consistently communicate your limitations, thoughts and decisions on this topic over the past months and your legal chair now in several places displaying pro-AI-generated text stances and showing very little regard or respect for fanfiction writers led to the situation at hand. All we ask for is transparency, and a legal team whose chair doesn't flippantly dismiss writers' concerns as too emotional or standing in the way of diversifying generative text. Please follow this up with a proper, thought-out statement, and continuous updates, so we can make an informed decision about our presence on and dedication to AO3.
    70. Reply
  70. androgenius commented: Agreeing with most people here. This was the LEGAL CHAIR of OTW and this "update" feels lukewarm at best. Betsy Rosenblatt does not understand AI even a little bit and it shows. The idea alone that money would be made of the fanworks that we do not make money off of FOR VERY GOOD REASON is 100% antithetical to what AO3 represents and frankly would land AO3 in serious legal trouble. None of this is comforting.
    71. Reply
  71. Stavia Scott Grayson commented: This is not the reassurance you think it is. If this is how Rosenblatt thinks, get rid of her. Writers do not want AI scraping their fics. No donations will be made to your drives until this situation is a) remedied and b) clarified.
    72. Reply
  72. betsy is a cunt commented: this is a horse shit statement fuck you we want confirmation that your LEGAL CHAIR doesn’t speak for all of you and that ai will not be allowed, used on ao3
    73. Reply
    • DL commented: If there's a moderator, could this comment be removed? I'm unhappy with the OTW but could do without the sexist slurs, thanks.
      • Reply
    • Ring commented: The slur is completely uncalled-for.
      • Reply
  73. K commented: I am disappointed that you removed it without archiving or saving a copy, or editing in a disclaimer. As someone who did not read it before it was removed, I am deeply concerned that you simply hid the details from those of us who didn't read it right away, leaving us to guess from what others say what the details were. I'm unimpressed by what they appear to have been, but I am also upset that I cannot reference them.
    74. Reply
    • K commented: Found it on the wayback machine. Honestly? I thought I was going to read something way worse than I did. You should put a "this is what we removed if you want to see it" link with it, probably. I still don't approve. Fanworks are not themselves public works but belong to their authors and are shared. And they are themselves based on other properties which are VERY often not public works. In some case a fanwork may quote some of the words from that work (as when doing an outside POV of a scene) and/or make use of a particular character's or work's native idiom or tone. AI training usage of these works WILL infringe on the fanfic author but may also infringe on the owner of the original work it's based on, if some of their words are included in the training data without being paid. Which may, in turn, lead to fanfiction itself being less-welcome as it was in the past. This is the opposite of protecting and preserving fanworks. I realize that one person saying one thing is not representative of the organization, but her position (in the organization) and her position (about AI) is worrisome. I am sharing my fics with anyone who wants to read them, but I am not agreeing to have an algorithm incorporate them. I do not want that to change in the future. I also recognize you cannot stop an unethical AI company from scraping the site. But I want a commitment from your organization, that regardless of her opinion (or anyone's opinion), no *willing* AI collaboration will be rolled out without an opt-in system (so the author must flag the fic as being available for that purpose) at any point, and that any such cooperation or collaboration will be communicated in advance of being implemented.
      • Reply
  74. N3GatorFan commented: I personally find the ideas of AI-generated stories and fanfic-trained AI disturbing. However, it sounds like Ms. Rosenblatt and OTW are taking a laissez-faire approach to AI. The debate about its legality is coming at the same time SCOTUS is set to rule on the "fair use" doctrine and copyright infringement (Andy Warhol Foundation for the Visual Arts, Inc. v. Goldsmith). I'm assuming OTW's and Ms. Rosenblatt's logic is that the organization can't criticize AI-generated art if our own works could be deemed "legally problematic". It's 100% unfair to us fanfic writers and artists, but the alternatives are just as distasteful Link to an article from GBH about the SCOTUS case: https://www.wgbh.org/news/national-news/2022/10/12/the-supreme-court-meets-andy-warhol-prince-and-a-case-that-could-threaten-creativity
    75. Reply
    • Ring commented: If they had said that, I think it would have been fine overall and more people would have understood where they're coming from. I've seen conversations around this that are reasonable and understanding of how restricting AI fic on the archive would probably not be sustainable or good for the community overall. I can only speak for myself, but the tone of the comment is what troubled me, and the fact that she either didn't know or didn't acknowledge the extent of the issues people have with AI. If she didn't know...well, IMO there's no excuse for not knowing when she's working with an org that exists to protect creatives, but it happens. If she did know, she chose a completely unserious approach that treated people's fears as silly, which...again, if she's paying any attention at all, she should know that people are concerned about large language models being used to wipe out entire career paths. This is extremely common among people who are way into this tech, just like it was with NFTs; there is no way to take concerns seriously without admitting that full-speed development and adoption is going to hurt people, and full-speed development and adoption is where the money comes from, so you try to capture the opinions of people who haven't heard the criticism yet by ignoring the crit and talking up happy, harmless possibilities. The idea that fanfic could be a driving force in making LLMs less biased compared to the volume of material they've been trained on is fucking ludicrous, especially since plenty of fanfic is dripping with racist, sexist, heteronormative and cisnormative bias, but the tell here is that she specifically mentioned--unprompted--that she thinks regulation could prevent LLMs from having a diverse data pool if it limits scraping. That is not the neutral opinion of someone who just thinks they're cool. That is an AI maximalist position: either unregulated scraping of public data has to continue or something bad will happen. People who have invested money or reputation in LLMs are racing the clock to get this tech so fully embedded in online infrastructure and people's lives that the US government will find it politically poisonous to regulate. I always feel like a conspiracy theorist talking about this stuff because everything having to do with tech right now sounds like it'd get tossed out of a movie script for being too ridiculous, and I actually doubt that, if I'm right, Rosenblatt has any malicious intentions. But I cannot overemphasize how likely I think it is based on her comments that she genuinely believes LLMs need to be allowed to scrape diverse works to "learn" progressive values, and that she will continue to use her platform to try to make the case that this is a good thing unless someone is able to step in and tell her to stop. I don't necessarily think she should lose her position on the board, and that may not even be feasible; I DO think she probably needs to be reminded that if fandom largely does not believe a pro-AI stance is in its best interest, it is not up to her to make a case on our behalf arguing otherwise.
      • Reply
      • N3GatorFan commented: I agree with you on the lucrative potential to have unproblematic AI results and the dismissal of our concerns. At the same time, I can't help but to think of the parallels between the current debate and the legality of fanfic when I posted my first one--on another site--in 2015. The majority of the results then seemed to indicate that we were in danger of violating copyright because we never obtained the permission of the creators and/or copyright holders. Even several disgruntled "fans" within a fandom I'm in took to social media and claimed that our works should be taken down because they were copyright violations. (The fandom isn't an overly popular one, and there was a fandom war going on at the time, FYI.) I finally decided to go for it when I saw the show's creator practically bragging about us fanfic writers and fan artists. Fortunately, they're more acceptable now, and I am glad I made the decision to start posting my own fanworks. With the SCOTUS court case, I don't trust this particular court. I fear that SCOTUS might rule that using a copyrighted visual artwork to create a new work isn't transformative, and all works created in that manner are subject to litigation. As a result, I've practically stopped work on three WIPs and cross-posting stories from another site to AO3 until after the ruling. If the justices rule in our favor, I'll probably pick them up again and apologize to readers for the wait. If not, then, as much as I hate to do it, I'll take down my work to comply with the new legal doctrine.
        • Reply
  75. H. commented: That's legit absurd that you claim that it was just one of 900+ volunteers when it was the Legal Chair(!), the head of the department that's supposed to be protecting the fic through legal means. It's shouldn't tolerateded when she made herself explicitly clear about her support to tech corporations and against actual creators during a writers strike. Such person should NEVER be allowed anywhere close to fanworks. But it's already telling that Ao3 considers bot generated text fanfic when such belief is alrrady actively harming fandoms by allowing bots to replace actual fanwork creators.
    76. Reply
  76. Jane commented: Well dang, I'd like to throw my hat in the ring for Legal Chair, then, since that's apparently a non-consequential volunteer position who gets to chill with government officials while we chat about important federal policy.
    77. Reply
  77. Unicron commented: So instead of actually listening to people you just removed the part that we were complaining about. I bluntly feel that there needs to be a code of conduct for all volunteers regarding conflicts of interest. Betsy Rosenblatt's job means she is in favor of using AO3 to make a profit via AI, and that means she has no place in the organization if you ask me. Also there is a difference between a random tag wrangler having a bad take on a personal blog and the head of legal being given a platform via AO3 supporting something that will harm creators. As is I am done donating to OTW until you adopt stances against the use of AI and scrapers on the archive. I will donate to other nonprofits that actually have codes of ethics in place instead
    78. Reply
  78. Not A "Fan" commented: Betsy has already done a great deal of harm to fans by voicing the exact opposite of their wishes to the *US Copyright Office*. I have never seen a single positive word about AI fanwork generation from a fan. This REEKS of a payoff by AI corporations. I will be withdrawing my support as I refuse to fund a representative who ignores the extremely apparent wishes of those they represent - Betsy can fund her new AI archive with her new AI cash.
    79. Reply
  79. Kedreeva commented: Not thrilled to have seen AO3/OTW speak in favor of AI being used for "Creative" anything. AO3 isn't the "archive of AI's own" it's the archive of OUR own. OUR own. Us. Humans. Fans. AI really has no place in that, because AI doesn't belong to our group. It's not one of us. It can't be. Whether it's legal or not, or moral or not, doesn't actually matter. /AI is not part of 'our' own/ and what it produces doesn't belong here.
    80. Reply
    • Kedreeva commented: Coming back to add that if it's going to be allowed under the premise that "well it will still be posted but untagged" then imo it should be a new major archive warning tick box that's separate from the rest of them, so no one can wiggle out of with "choose not to warn" and it needs to be in the header, something people can easily filter against. Art AI has been having damaging real world consequences for artists, and now it's doing the same for writers, and none of that is going to get better. Making it mandatory to declare use of AI and making its use obvious and easily avoidable is the least you can do to help us.
      • Reply

Leave a Reply

Your email address will not be published. Required fields are marked *