Timeline for How do staff officially reconcile AI Assist with the reasons of the generative AI ban, as given in Stack Overflow policy?

Current License: CC BY-SA 4.0

Post Revisions

55 events

when toggle format	what		by	license	comment
Jan 26 at 23:14	answer	added	Ben Voigt		timeline score: 5
Dec 28, 2025 at 1:34	comment	added	Vince		The fact that this post, and all comments & answers that agree with it are being heavily downvoted is troubling to say the least. StackOverflow is about expert programmers donating their time to help each other. "AI Assist" is a great way to replace a valuable community of experts with an LLM, which is trained using the free labor of those very experts. And anyone who disagrees with the idea that this will decrease engagement with human-written questions & answers is lying at worst, and fooling themselves at best.
Dec 10, 2025 at 15:50	comment	added	Jane		I am also confused, I also felt it was contra to what I thought
Dec 8, 2025 at 15:08	comment	added	ygtozc		@MarkAmery I agree with Weijun Zhou's explanation. A grey area is inevitable here, unfortunately: Whether we like it or not, people here can and will often consult AI output before asking a question and it's not possible to evaluate how this influences questions objectively. However, I feel you are right to point out that AI Assist as a tool encourages people to post questions (mis)guided by LLM output, and that this seems against the motivations of the ban, if the intent is to reduce the influence of inaccurate and/or outdated genAI output on SO content.
Dec 8, 2025 at 14:23	comment	added	ygtozc		@Dominique No, I haven't used any AI tools at any point in writing this question. I avoid using so-called "generative AI" whenever possible, and I do not need it to research or to write. The similarity probably exists because AI output is tailored to resemble (formal, technical) writing from sites like Stack Overflow.
Dec 8, 2025 at 12:22	comment	added	Weijun Zhou		@MarkAmery I think the spirit of the linked post is that if you are not already in a good position to answer a question, using GenAI cannot help and using GenAI to help you craft an answer that you are not otherwise able to write is not acceptable. I don't think in terms of spirit it prevents necessary research before you ask a question and I think it is only natural to check for GenAI (that is, if you believe GenAIs are at least somewhat useful) as part of your research effort before asking a question.
Dec 8, 2025 at 8:03	comment	added	Dominique		Please don't kill me for asking this, but the format of this question looks quite familiar. Did you use A.I. to write this question? 😊
Dec 7, 2025 at 8:38	comment	added	Mark Amery		I'm surprised so little has been made of the conflict with the stricter interpretations of the GenAI ban. A significant slice of the Meta community and mod team have taken the view that we do not merely ban copying & pasting or close-paraphrasing GenAI output, but also any use of GenAI to aid your posting - even pure research use. Hundreds of posts were deleted on that basis. But using AI Assist and then asking a question when it suggests that you do so is precisely that - research use to aid posting, which many have always said is banned.
Dec 7, 2025 at 2:31	history	edited	ygtozc		add ai-generated-content tag
Dec 7, 2025 at 2:02	comment	added	ygtozc		@philipxy Even if your concern is solely with my usage of the word "policy", the question you refer to still frames the reason for the ban in terms of the unreliability of generative AI technologies at large. It is still factually correct, as written in the question, that the presence of AI Assist runs against the motivations and stances cited in a question post that has been explicitly labeled "Policy". If you have a better term to suggest here, e.g. "Help Centre information" or "SO Help", to refer explicitly to Help Centre content, please inform me and I will edit the question.
Dec 7, 2025 at 1:55	comment	added	ygtozc		@philipxy Are statements in the SO Help Centre not considered part of SO policy? I apologize if the term "policy" strictly refers to the kind of post you linked to, please tell me if there is a better term to describe content written on the SO Help Centre to presumably describe the website's stances on important matters. I think it is reasonable to refer to an SO Help Centre article, and motivation and stances expressed therein, as "SO policy", particularly when the policy you cited directs the reader to that exact Help Centre page for further information.
Dec 7, 2025 at 1:13	comment	added	philipxy		The link on AI policy is clear that it is about posting & not about use of AI in other aspects of the site. You have no reason to claim "contradictions" by staff. You continue to be oblivious to what you have been told.
Dec 7, 2025 at 0:48	comment	added	ygtozc		@philipxy Staff behavior contradicting (the motivations stated in) a piece of policy is not resolved by citing the policy itself. This is discussed at length in this question. I am not sure how I could be clearer on this point within the question: The policy says "LLMs are bad", SO staff says "LLMs are good". To add, the question you linked is not about AI Assist in specific, which is central to my question. The linked question is also too old to accurately reflect current staff policy re: AI Assist.
Dec 6, 2025 at 16:48	review	Close votes
Dec 8, 2025 at 17:27
Dec 6, 2025 at 16:05	comment	added	philipxy		This question is similar to: Policy: Generative AI (e.g., ChatGPT) is banned. If you believe it’s different, please edit the question, make it clear how it’s different and/or how the answers on that question are not helpful for your problem.
Dec 6, 2025 at 1:31	answer	added	n00dles		timeline score: -11
Dec 5, 2025 at 12:09	comment	added	ygtozc		@philipxy Thank you for the suggestions, I genuinely appreciate them. This is the first time I have had a question closed on Stack Exchange, so I was not sure what the best way to address these changes would be. I'll follow your suggestions in the future.
Dec 5, 2025 at 9:30	comment	added	Lundin		Wikipedia article of the day: Doublespeak.
Dec 5, 2025 at 2:10	comment	added	philipxy		Please avoid meta commentary in posts. Don't tell us about your edits, just edit the post to be the best presentation possible. Don't write "to clarify", clarify. Don't add "clarification", rewrite what isn't clear. PS There are no answers, you might consider posting a new question if you think because of its previous assumptions you are essentially proposing a new discussion topic. PS Clarify posts via edits, not comments; delete & flag obsolete comments.
Dec 4, 2025 at 20:58	comment	added	user400654		Probably the same way they justified adding the ai assist link to the left column across the network without prior notification as agreed to after the mod strike: “technically… we aren’t adding anything new to the network, it’s just a link going somewhere else!” In this case, technically it isn’t posting answers to questions in Q&A…
Dec 4, 2025 at 20:32	history	edited	cafce25	CC BY-SA 4.0	Reorder meta commentary to the end as it's relevant to the comments, but otherwise not really necessary for the question as it stands now.
Dec 4, 2025 at 20:27	history	reopened	il_raffa Karl Knechtel discussion
Dec 4, 2025 at 18:13	comment	added	ygtozc		I've edited the question and requested it to be reopened. I hope it is now much clearer, I have decided to reframe it as a question of how SO staff reconcile AI Assist with community policy. That said, I do not have high expectations of this question receiving a meaningful answer.
S Dec 4, 2025 at 18:07	review	Reopen votes
Dec 4, 2025 at 20:34
S Dec 4, 2025 at 18:07	history	edited	ygtozc	CC BY-SA 4.0	Removed all discussion of ban violations, focusing instead on contradictions between AI Assist and motivations/values stated in SO policy. Added to review
Dec 4, 2025 at 1:46	comment	added	Weijun Zhou		The question may be better received if it focuses on the objective contradiction mentioned in the comments instead of based on a false premise ("clearly lifted"). That said I don't see how such a discussion will change the current status as neither the company nor the community is willing to step back from what they insist on.
Dec 4, 2025 at 0:26	comment	added	ygtozc		@SecurityHound Thank you for informing me. I saw some discussion of enforceability today, but didn't read enough to add it to this question, so this is useful. user400654, that's really unfortunate to hear, but I am glad I am not the only one who's not happy
Dec 4, 2025 at 0:18	comment	added	Security Hound		@ygtozc - The community and community moderators delete AI content. SE staff do not delete AI content. SE staff just agreed that the policy could be enforced, after I believe saying, it could not be enforced.
Dec 4, 2025 at 0:05	comment	added	user400654		I mean.. there’s other places frustration can be expressed, but none of them are any more effective than this comment section 🤷‍♂️we’ve been fighting this fight for a long time now, and from my perspective, I see no signs of the slide stopping.
Dec 3, 2025 at 23:56	comment	added	ygtozc		@user400654 Thank you, it's relieving to hear that. "They can implement features that help people violate policy" at least makes me feel the problem isn't in my head. Do you know of a better place or way (if any exists) I could express my concerns? If the answer here to my questions is as simple as "Stack the company and Stack the community can contradict each other because the former does what it wants", that resolves my complaint about the policy being outdated.
Dec 3, 2025 at 23:55	comment	added	cafce25		Not at all, the marketing claims the company makes about AI assist are bold, but that doesn't mean attribution works flawlessly or that the answers it generates are always 100% reliable without checking.
Dec 3, 2025 at 23:53	comment	added	ygtozc		In other words, the rules of the policy might not be outdated, but the motivations of the policy definitely are. Again, the Help Center currently says "There's no point in us serving an LLM when anyone could just ask another LLM"... while serving an LLM. Maybe I'm pedantic, but this makes it hard to trust site policy. SO staff is either marketing a feature they think/know is unreliable, or they do not really believe the principles they espouse in the policy. This would be resolved by (edit) justifying how AI answers that aren't accurate or human enough to post are good enough for new users
Dec 3, 2025 at 23:51	comment	added	user400654		Just as it was their choice, to not touch the network feature-wise for 10 years while it slowly declined, year after year, due to various reasons. :shrug: they decided their SaaS product was more important
Dec 3, 2025 at 23:51	comment	added	cafce25		SO is supposed to be a library of Q/A. as such the quality bar for content posted here is extremely high. That doesn't mean that content that doesn't pass that bar can't be useful, but not every Q/A belongs in a library of programming problems.
Dec 3, 2025 at 23:45	comment	added	user400654		See… the thing is, this policy is community generated and community enforced. Stack can, if they so choose, put an end to it… but they haven’t. They can also choose to implement features that help people violate the policy, which they have in the past, and quickly rescinded due to community pushback. It’s their platform, it is within their right to run away their community in service of their AI ambitions.
Dec 3, 2025 at 23:45	comment	added	ygtozc		@cafce25 "Do not use AI Assist to post" is clear. "AI assist is good" is clear. "AI Assist is good enough to use but not good enough to post" is not clear, because the AI Assist is presented as simultaneously good enough to consume responses but not good enough to generate them. If AI Assist makes SO responses more accessible (that is the claim), then LLMs can provide reliable and credited responses and we have one that does. If that is SO's stance, then the policy can no longer claim LLMs are (AI assist is) banned because they are unreliable, as it currently explicitly does in Help Center
Dec 3, 2025 at 23:34	comment	added	ygtozc		@user400654 I see your point. There's something I have a lot of trouble reconciling with the policy though, and that's saying "LLMs cannot be trusted, so you cannot use even ours to post" and "our LLM is so great that it'll give you correct information and credit the correct sources" at the same time. If LLMs are banned because of XYZ, and AI Assist fixes XYZ, why is AI Assist banned from SO like every other LLM? Conversely, AI Assist does not give human responses, so it runs against the Help Desk's policy claims of centering human answers
Dec 3, 2025 at 23:34	comment	added	cafce25		"how to use AI Assist without violating current SO policy in general?" is extremely trivial, you can use it however you want as long as it's not for creating content on SO. The majority of intended usage scenarios of SO do not involve any content creation.
Dec 3, 2025 at 23:18	comment	added	ygtozc		Am I wrong in believing that Stack Overflow has a responsibility to publish extremely clear policy around 1) How they will prevent these incentives provided by AI Assist from compromising human responses and 2) how to use AI Assist without violating current SO policy in general? Even if we put aside all compliance and quality concerns, just practically, isn't it going to be a nightmare to moderate all accidental infractions when it's so easy to miss this nuance?
Dec 3, 2025 at 23:11	comment	added	user400654		@ygtozc this feature was first introduced in July and has coexisted for the entire period of time since then to now without much conflict in terms of this policy, I don’t see what embedding it on the home page changes about the policy and our ability to enforce it. I definitely agree that it is a step in the wrong direction if we intend to increase participation, but that’s another matter.
Dec 3, 2025 at 23:09	comment	added	ygtozc		It is practically impossible to enforce "do not use LLMs to post here because they suck" while simultaneously advertising an LLM that, according to you, does not suck. Edit: On top of that, you aren't teaching people to read more answers by summarizing them, you're encouraging them to NOT engage with anything human-written unless they have to. AI assist therefore 1) means more people will post AI replies whether that's policy or not, and 2) fewer people will go out of their comfort zone to find and engage with replies if they can just get what info they need and leave.
Dec 3, 2025 at 23:08	comment	added	ygtozc		@SecurityHound You have touched upon something that frustrates me greatly. Arguably my biggest complaint is this: If you incorporate an LLM into your website, people can and will use it to generate unverified responses as answers, because it's right there. No matter how carefully policies are worded, people will still be encouraged to use it for that purpose, because it makes no sense to claim that LLMs are forbidden for being unreliable while also saying AI assist is a worthwhile tool because it can provide reliable answers, which are even CREDITED reliably.
Dec 3, 2025 at 23:04	comment	added	ygtozc		@cafce25 As quoted above, the ban 'includes "asking" the question to an AI generator then copy-pasting its output'. I interpreted this as "you cannot post questions that are just requesting an explanation of AI output". I now realize this might've been a major mistake - reflecting more on it, it's more likely that this means "when ANSWERING a question, you cannot just copy over AI responses". I was thrown off by the phrasing - "when posting content in Stack Overflow" implies BOTH questions and answers, to me, so I thought AI output in questions was also forbidden (unless it was detailed).
Dec 3, 2025 at 23:00	comment	added	Security Hound		@ygtozc - The policy with regards to AI generating (questions and answers) has not changed regardless if the AI Assist feature exists on the network. If anything you can use AI Assist to everything except generate that content on your behalf. If AI content is allowed on this network, that is the end of the network for many users, since there is no quality control on AI generated content by those users who have or will use it.
Dec 3, 2025 at 22:55	comment	added	ygtozc		@starball I agree that clarifications to policy would be useful (if not necessary). Would that be sufficient grounds for reopening the question (after removing the assertion that AI Assist represents a lift of the ban)? Providing an AI tool without explaining how it relates to a general AI ban "when posting" seems like a recipe for confusion, to me. I also still believe there are contradictions in how LLMs are described in policy and AI Assist - it doesn't add up to say LLMs cannot provide attribution or reliable info (in Help Centre) and to say AI Assist does exactly that, for instance.
Dec 3, 2025 at 22:55	comment	added	cafce25		@ygtozc why would that be not allowed, you can't use GenAI to write a question, but you can absolutely ask about AI generated code
Dec 3, 2025 at 22:49	comment	added	ygtozc		I see your points, thank you both. I understand that AI Assist might coexist with this ban under such an interpretation. However, I believe there are still way too many ambiguities in the policy as written regarding how AI Assist fits into this policy. At a minimum, if AI Assist can provide answers and tell users to post questions, how this can be done under a generative AI ban needs to be clearly explained as site policy. People can and will post questions to the effect of "This is what AI Assist told me about X, can someone explain?" - this is explicitly forbidden under current policy.
Dec 3, 2025 at 22:32	comment	added	starball Mod		and the big headline on the meta post, that says "All use of generative AI (e.g., ChatGPT and other LLMs) is banned when posting content on Stack Overflow." (italics added). Maybe these writings could be clarified so people understand that it's not about "AI Assist", but as far as I know, the policy hasn't changed.
Dec 3, 2025 at 22:32	comment	added	starball Mod		My understanding of the scope of the GenAI policy is that it's about what you post. (If I'm wrong on this, I'm sure someone will quickly correct me). My cues are the historical context, phrasing like "content for Stack Overflow", and "Posting content generated by generative artificial intelligence tools may lead to a warning from moderators, or possibly a suspension for repeated infractions.",
Dec 3, 2025 at 22:30	comment	added	NoDataDumpNoContribution		AI technology can be freely used by the company in any way they see fit. For a time they actually really posted LLM generated content as answers earlier this year. However, users are forbidden to use AI generated content when posting. It was always like this. But I see the point that this might seem a bit unfair.
Dec 3, 2025 at 22:30	history	closed	MakyenMod		Duplicate of Policy: Generative AI (e.g., ChatGPT) is banned
Dec 3, 2025 at 22:29	comment	added	user400654		@ygtizc thus far, this feature does not assist people in creating content for posting on SO. It already has the feature they mentioned that you specifically called out that creates a pathway to asking on SO… it’s just a link to the ask wizard.
Dec 3, 2025 at 22:28	comment	added	ygtozc		@user400654 Could you explain how a generative AI ban can coexist with a generative AI feature that is intended to explicitly guide how people contribute to the site?
Dec 3, 2025 at 22:22	comment	added	user400654		It hasn’t been lifted. Nothing has changed.
Dec 3, 2025 at 22:14	history	asked	ygtozc	CC BY-SA 4.0

toggle format