User talk:Chaotic Enby

Welcome to Meta!

Hello, Chaotic Enby. Welcome to the Wikimedia Meta-Wiki! This website is for coordinating and discussing all Wikimedia projects. You may find it useful to read our policy page. If you are interested in doing translations, visit Meta:Babylon. You can also leave a note on Meta:Babel or Wikimedia Forum if you need help with something (please read the instructions at the top of the page before posting there). Also worthwhile acquainting yourself with the functions of global user pages. Happy editing!

Queen of Hearts (talk) 19:33, 12 March 2024 (UTC)

Reminder to vote now to ratify the Wikimedia Movement Charter

You can find this message translated into additional languages on Meta-wiki. Please help translate to your language

Dear Wikimedian,

You are receiving this message because you previously voted in the 2021 Movement Charter Drafting Committee (MCDC) election.

This is a reminder that if you have not voted yet on the ratification of the final Wikimedia Movement Charter draft, please do so by July 9, 2024 at 23:59 UTC.

You can read the final text of the Wikimedia Movement Charter in your language. Following that, check on whether you are eligible to vote. If you are eligible, cast your vote on SecurePoll.

On behalf of the Charter Electoral Commission,

RamzyM (WMF) 15:24, 5 July 2024 (UTC)

Wikisporadic Newsletter: September 2025

Wikispore gets shout-out on list of "Ideas originated from Wikimania" this past month in Nairobi, Kenya.

Wikisporadic, an Occasional Newsletter for our Germinating Community.

• Sister Projects Task Force and RFC After the recent consultation about Wikispore, an RFC has been launched on Requests for comment/Sister Projects next steps#Approve Wikispore as Incubator-like project? and related questions. See also Wikinews Pulse proposal.
• Canadian art project Parallel: The History and Archives of Artist-Run Centres in Canada (Q135767262) has been funded and will publish material to wikispore:Art Spore, alongside other GLAM-type initiatives. See also Wikispore Basic proposal.
• Wikimania Watch Keep Wiki Weird session in Nairobi, as well as Wikispore meetup and a continuing regional interest in better representing oral knowledge on-wiki.
• Next strategic call After Wikispore Day celebrations in July and August, help pick a date for our follow-up September Strategy Spectacular for the first week of the month.

Participate: Comment on stories and ideas in this Newsletter, Join Wikispore on Telegram, Join Wikispore mailing list --Pharos via MediaWiki message delivery (talk) 14:50, 28 August 2025 (UTC)

Thanks @Pharos! Regarding the September meeting, do you think it would be worth it to have a link to it on the Wikispore main page, or maybe a small calendar of future events? The project is still quite small, and it could help with organizing (edit: moving the conversation to Talk:Wikispore which is more visible). Chaotic Enby (talk) 17:40, 28 August 2025 (UTC)

Reminder: global rename policy requirement

You are receiving this message because you are either a Steward or a global renamer. If you are steward and don't perform global rename, you can safely ignore this message.

Hello, this is a general reminder that renames or vanish requests should be reserved for someone who knows the language and local wiki affairs on their home wiki or most active wiki for the first 24 hours after the request is placed.

For the first 24 hours after a rename or vanishing request has been placed, it should only be processed by a renamer who has sufficient proficiency in the language of the requester's home or most active wiki, and is active on that wiki.

— Global renamers#Guidelines

Refer to the 2025 RFC adding this rule, and following global-renamers mailing list emails: 1, 2, 3. Thank you.

— regards, Revi delivered by MediaWiki message delivery (talk) 07:11, 8 April 2026 (UTC)

Discussions

I notice that you've started two global RFCs in the last two months. Have you been discussing these proposals somewhere in advance? In the global community, bold proposals are often not successful. WhatamIdoing (talk) 03:05, 26 April 2026 (UTC)

Hi! The previous proposal was discussed for a long time at en:Wikipedia:Village pump (proposals)/Archive 227#Request for community_ consensus – CentralNotice for Wiki Loves Ramadan_2026, where there was support for bringing it to Meta for a wider discussion, which I subsequently did. Regarding the latest proposal, I workshopped it with other editors (most notably Aca), although in retrospect greater participation would have been helpful. The main issue is that it is difficult to get a wide audience on Meta without directly starting an RfC or using something like a mass message notice, which can easily appear disruptive if it is only for informal workshopping. Do you have any advice on how to proceed? Chaotic Enby (talk) 11:24, 29 April 2026 (UTC)

I've not found that enwiki's view necessarily translates to global support. The problems enwiki deals with are different from most other communities (e.g., enwiki gets vastly more paid editors).

The ideal approach is that all the stakeholders already know about it when the proposal is posted. A chat at Meta:Babel or a relevant page might help. Connecting with another big wiki (e.g., asking for the frwiki POV at w:fr:WP:BISTRO), a chapter, or a relevant group here at Meta-Wiki is also good. For example, the Small Wiki Monitoring Team might be able to give you useful information about how AI is affecting small wikis, but the act of asking them is going to let them know that this is something that concerns you. The process isn't quick, but it is usually effective in the end. WhatamIdoing (talk) 19:03, 29 April 2026 (UTC)

Thanks for the advice! Chaotic Enby (talk) 23:30, 29 April 2026 (UTC)

AI/Draft Policy= my view

Definitions

Content

The reader-facing aspects of the project. This may be encyclopedia articles on Wikipedia editions, books on Wikibooks editions, items on Wikidata, media on Wikimedia Commons etc.

Generating

Creating original content; whether from an existing basis or from scratch. This excludes importing existing content near-verbatim (e.g. through OCR), but includes creative modifications of such content.

Disclosure

Indicating that AI assistance has been used and to what extent, either on the page itself, in an edit summary, on the talk page, or through any other channel deemed suitable by the project. Specifics such as models or prompts are not required for disclosure.

Basic copyediting

Any formatting of content that does not change its underlying meaning. For instance, this applies to fixing grammar or orthography, formatting lists into tables, or formatting a citation template. Creative interpretations of existing content are excluded from this definition.

Human review

A thorough reading and editing of the AI-generated content, including verification that any generated citations exist and accurately support the corresponding content.

Opt-out policy

A policy that applies to Wikimedia projects by default, but can be rescinded or amended through local consensus. It will not apply by default to any project that, in the past, formally discussed a policy or guideline regarding generative AI, whether successfully or unsuccessfully.

Artificial intelligence (AI)

Refers to any computational system capable of generating, transforming, or analysing text, media, or data in ways that resemble human reasoning or language production. This includes large language models, machine‑translation systems, automated summarisation tools, and any software that produces or meaningfully alters content without direct human authorship.

Translation

Refers to the process of converting text from one language into another while preserving meaning, tone, and factual accuracy. For the purposes of this policy, translation is considered a mechanical linguistic transformation rather than a substantive content change, provided the editor verifies the accuracy of the translation and ensures that no new information, claims, or interpretations are introduced.

Grammar correction

Refers to the improvement of spelling, punctuation, syntax, and basic sentence structure without altering the factual content, meaning, or intent of the text.

Language refinement

Refers to minor improvements in clarity, readability, or fluency that do not change the underlying meaning, introduce new claims, or alter the editorial stance of the text. Examples include smoothing awkward phrasing, improving flow, or adjusting tone for consistency.

Content

Content, for the purposes of this policy, refers to the reader‑facing outputs of Wikimedia projects, including encyclopedia articles on Wikipedia editions, books and instructional materials on Wikibooks, structured data items on Wikidata, media files and descriptions on Wikimedia Commons, and similar materials across all other projects. Because this content is what readers directly encounter, any use of artificial intelligence in its creation or modification must be handled with exceptional care, transparency, and accountability.

• AI may be used to support content development, but editors must disclose AI involvement whenever the tool contributes to the substance, structure, or meaning of the material. Disclosure must appear in edit summaries or on talk pages in a manner that is visible and understandable to other contributors. The only exceptions to this requirement are when AI is used exclusively for translation, grammar correction, or language refinement, provided that these uses remain strictly mechanical and do not introduce new claims, alter meaning, or affect the informational substance of the content. These exceptions recognise that certain linguistic transformations do not meaningfully change the underlying content and therefore do not require the same level of transparency.
• All AI-generated or AI-assisted content must undergo human review before being published or left in place. Human review is not a formality; it requires the reviewing editor to read, understand, and critically evaluate the material to ensure that it complies with core Wikimedia policies, including verifiability, neutrality, and no original research. The reviewer must confirm that all claims are supported by reliable sources, that citations are valid and not fabricated, and that the content does not contain hallucinated details, biased phrasing, or misinterpretations of complex topics. Editors remain fully responsible for any errors introduced through AI use, and AI cannot be cited as justification for policy violations.
• AI must not be used to generate unsourced claims, invent citations, or create large volumes of new content without prior community approval. Projects may establish additional safeguards for sensitive areas such as biographies of living persons, medical information, legal topics, or rapidly evolving current events, where AI-generated content poses heightened risks. In such areas, communities may require stricter review standards or prohibit AI-generated contributions entirely.
• In discussions—including Requests for Comment, noticeboards, talk pages, and other community deliberative spaces—AI may be used to assist in drafting comments, but editors must disclose this use whenever the AI meaningfully shapes the substance of their contribution. AI-generated arguments must not be presented as personal opinions, and AI must not be used to simulate consensus, generate mass comments, or influence discussions in ways that obscure genuine community participation.
• Limited, non‑substantive uses such as grammar correction or minor language refinement in discussion posts do not require disclosure.
• AI tools may support moderation and administrative workflows, such as identifying vandalism, suggesting actions, or flagging problematic edits, but they must not take autonomous actions such as blocking users, deleting pages, or closing discussions. Human judgment is required for all final decisions.
• Communities should be informed when AI-assisted moderation tools are in use, and such tools must provide transparent reasoning that can be reviewed by human editors.

This baseline policy applies to all Wikimedia projects except those that qualify for and exercise an opt‑out under the provisions below. Projects may adopt stricter rules but may not operate below these minimum standards.

Opt-out policy

An opt‑out policy is a global rule that applies to Wikimedia projects by default but may be rescinded or amended through local consensus. Under this framework, any Wikimedia project may choose to restrict or prohibit the use of AI-generated or AI-assisted content, including within Content and community discussions, provided that the decision is made through a clear, well-documented local process. A project that opts out may forbid AI from generating or expanding articles, books, data items, media descriptions, or other reader‑facing outputs, and may also restrict AI use in discussions if the community believes it undermines authentic participation or introduces risks inconsistent with local norms.

Crucially, this global policy does not apply by default to any Wikimedia project that has previously held a formal discussion—successful or unsuccessful—regarding generative AI policies or guidelines. Any project that has already debated AI use retains full autonomy over its own approach and is not automatically bound by the global baseline. This respects the principle of local self-governance and acknowledges that many communities have already developed their own frameworks, expectations, or norms around AI use.

• Projects choosing to opt out must document their decision on a publicly accessible policy or guideline page, specifying the scope of the restriction, the types of AI use that are prohibited or permitted, and any exceptions they choose to allow. They may also establish local enforcement mechanisms, such as reverting AI-generated edits, issuing warnings, or restricting the use of AI tools by specific users.
• Opt‑out communities may still permit limited uses of AI—such as grammar correction, translation, or minor language refinement—if they choose, but they are not required to do so.
• Even when opting out, communities must continue to uphold global expectations of transparency and accountability. Editors who use AI in violation of a local opt‑out rule must still disclose that use, and undisclosed AI involvement may be treated as a policy breach.
• Communities may revisit, revise, or reverse their opt‑out decisions at any time through local consensus, ensuring that their approach to AI remains responsive to evolving needs and circumstances.

Human review mechanisms

• The reviewer must verify that all factual claims are supported by reliable sources, that citations are valid and not fabricated, and that the material complies with core policies such as verifiability, neutrality, and no original research.
• This includes checking for subtle issues that AI systems commonly introduce, such as confidently stated inaccuracies, misinterpretations of sources, biased phrasing, or invented details.
• The reviewer must also ensure that the tone, structure, and framing of the content align with established community norms and do not introduce undue weight or misleading emphasis.
• Human review must involve genuine editorial judgment. The reviewer should compare the AI output against source material when appropriate, confirm that quotations and paraphrases are accurate, and ensure that the content does not distort or oversimplify complex topics.
• When reviewing translations or language refinements, the editor must confirm that the meaning has not shifted and that no new claims have been introduced. If the content concerns sensitive areas—such as biographies of living persons, medical information, legal topics, or current events—the reviewer must apply heightened scrutiny and may need to consult additional sources or subject‑matter experts.
• A valid human review cannot be superficial. Simply skimming the text or assuming that the AI is correct is insufficient.
• The reviewer must be confident enough in the subject matter to detect errors and must be willing to revise, correct, or remove AI‑generated material when necessary.
• If the reviewer cannot verify the accuracy of the content, they should not approve it.
• Once the review is complete, the human editor assumes full responsibility for the final version of the content, including any errors that remain.

Please see whether this is an effective measure: It took a while to write this up...

Thanks, Globetrotter30 (talk) 08:48, 16 May 2026 (UTC)