Platform Policy Watch

← All pages

X's Privacy Policy contains two distinct, consistently-present disclosures about AI training. First, X states it may use information users provide and "publicly available information" — explicitly including public posts — to train its own machine learning and AI models (Section 2.1, "Operate, improve, and personalize our services"). Second, under "Third-party collaborators," X discloses that when it shares or discloses user data with third parties (per the user's sharing settings), those third parties may independently use that data "to train their artificial intelligence models, whether generative or otherwise" — a use X says is outside its own control. Both clauses are present, word-for-word identical, in every single scrape from 2024-10-17 through 2026-01-16 — this is a long-standing, stable policy, not a recent addition.

What this wiki found — complete, every page

AI trainingRoyalty-freeSublicensablePerpetual / irrevocableTransferableDerivative works

Presence facts, not verdicts — each flagged term links to its definition and the exact clause on this page. Absence of a badge means the term isn't currently flagged here, not that the page is risk-free.

Independent reviews, where they had something to say

X – Privacy Policy

Dataset: PGAv2 (Platform Governance Archive v2) — 25 platforms, 2022–2026

PGAv2 (Platform Governance Archive v2) is this wiki’s legacy dataset: pre-tagged JSONL records covering 25 major social-media, messaging, and content-sharing platforms, with risk scores and keywords assigned via systematic extraction of 1,736 high-risk clauses from each platform’s Terms of Service, Privacy Policies, and Community Guidelines, spanning 2022–2026.

Note: This page contains documented policy clauses. Risk assessment is qualitative and context-dependent. For analysis of patterns across platforms, see:

Source: sources/jsonl/X_Privacy_Policy.jsonl Last updated: 2026-01-16

Overview

X’s Privacy Policy contains two distinct, consistently-present disclosures about AI training. First, X states it may use information users provide and “publicly available information” — explicitly including public posts — to train its own machine learning and AI models (Section 2.1, “Operate, improve, and personalize our services”). Second, under “Third-party collaborators,” X discloses that when it shares or discloses user data with third parties (per the user’s sharing settings), those third parties may independently use that data “to train their artificial intelligence models, whether generative or otherwise” — a use X says is outside its own control. Both clauses are present, word-for-word identical, in every single scrape from 2024-10-17 through 2026-01-16 — this is a long-standing, stable policy, not a recent addition.

The risk score in the raw data swings between 50 and 80–100 across scrapes, but this is not a sign of the policy itself changing. It reflects the scraper occasionally capturing one extra sentence of surrounding text (a “wider window”) around the same paragraph. The one real edit found — see Changes Summary — is a help-center link’s URL slug, updated from twitter-services-and-corporate-affiliates to x-services-and-corporate-affiliates sometime between the 2024-10-17 and 2025-06-28 scrapes, consistent with X’s ongoing Twitter→X rebrand cleanup. Notably, the 2024-10-17 scrape captured both the old and new link forms simultaneously — most likely from two different on-page renderings of the same “Learn about affiliates” link (e.g., a summary widget vs. the main policy body) that hadn’t yet been updated in sync.

Flagged Keywords & Risks

Legal Context & Research Significance

Under Atkinson (2025), a platform’s terms become enforceable against scraping bots only when those terms are themselves disclosed on the pages a bot accesses; X’s AI-training clause works in the opposite direction — rather than prohibiting third-party scraping for AI, it is X’s own first-party admission that it trains its “machine learning or artificial intelligence models” on public posts and user-provided data, a disclosure that has remained word-for-word identical across every scrape from 2024-10-17 through 2026-01-16. This clause falls squarely into wearetosed’s “data sharing & selling” and “tracking & profiling” categories, compounded by the third-party clause stating that recipients of shared data “may use it… to train their artificial intelligence models” outside X’s control. Per Soneji et al. (2025), clauses like this one typically favor the service provider in roughly two-thirds of ToS;DR-style cases, and the absence of any AI-specific opt-out in either snippet is consistent with that provider-favoring pattern even though the underlying text itself is not especially hard for a lay reader to understand.

Changes Summary

DateWhat changed
2024-10-17Baseline version (first scrape in this dataset) — core AI-training clause and third-party-AI-training clause both present. This scrape uniquely also captured a wider-window duplicate of the core clause still using the pre-rebrand twitter-services-and-corporate-affiliates help-center link, alongside the normal capture already using x-services-and-corporate-affiliates.
2024-11-16Wider-window duplicate (with the old twitter-services link) no longer captured; the standard x-services-link version of the core clause and the third-party clause are both unchanged.
2025-04-17No changes from previous version.
2025-05-09No changes from previous version.
2025-05-21No changes from previous version.
2025-06-28Wider-window duplicate of the core clause reappears, now using the updated x-services-and-corporate-affiliates link (rebrand fully applied); underlying clause text otherwise unchanged.
2025-07-29Wider-window duplicate no longer captured; core and third-party clauses unchanged.
2025-10-23No changes from previous version.
2025-12-17Wider-window duplicate of the core clause captured again — byte-identical to the 2025-06-28 capture; underlying clause text unchanged.
2026-01-16Wider-window duplicate no longer captured; core and third-party clauses unchanged.

Version History

2024-10-17

Clause: artificial intelligence, machine learning, train AI/models

For example, the same piece of information may be used differently for different purposes to ultimately deliver a single service. We think it’s most useful to describe the five main ways we use information and if you have questions that are not answered, you can always contact us. Here we go: 2.1 Operate, improve, and personalize our services. We use the information we collect to provide and operate X products and services. We also use the information we collect to improve and personalize our products and services so that you have a better experience on X, including by showing you more relevant content and ads, suggesting people and topics to follow, enabling and helping you discover affiliates, third-party apps, and services. We may use the information we collect and publicly available information to help train our machine learning or artificial intelligence models for the purposes outlined in this policy. We may use the information we collect from accounts of other services that you choose to connect to your X account to provide you features like cross-posting or cross-service authentication, and to operate our services. We use your contact information to help others find your account if your settings permit, including through third-party services and client applications. We use your information to provide our advertising and sponsored content services subject to your settings, which helps make ads on X more relevant to you. We also use this information to measure the effectiveness of ads and to help recognize your devices to serve you ads on and off of X.

Clause: artificial intelligence, train AI/models

When you watch or otherwise interact with content from our video or cross-platform sharing partners, they may receive and process your personal information as described in their privacy policies. For video content, you can adjust your autoplay settings if you prefer that content not to play automatically. Third-party collaborators. Depending on your settings, or if you decide to share your data, we may share or disclose your information with third parties. If you do not opt out, in some instances the recipients of the information may use it for their own independent purposes in addition to those stated in X’s Privacy Policy, including, for example, to train their artificial intelligence models, whether generative or otherwise. Through our APIs. We use technology like APIs and embeds to make public X information available to websites, apps, and others for their use, for example, displaying posts on a news website or analyzing what people say on X. We generally make this content available in limited quantities for free and charge licensing fees for large-scale access. We have standard terms that govern how this information can be used, and a compliance program to enforce these terms. But these individuals and companies are not affiliated with X, and their offerings may not reflect updates you make on X. For more information about how we make public data on X available to the world, visit https://developer.x.com.

Clause: artificial intelligence, machine learning, train AI/models (wider-window duplicate of the clause above — captured this scrape using the pre-rebrand twitter-services-and-corporate-affiliates help-center link)

We think it’s most useful to describe the five main ways we use information and if you have questions that are not answered, you can always contact us. Here we go: 2.1 Operate, improve, and personalize our services. We use the information we collect to provide and operate X products and services. We also use the information we collect to improve and personalize our products and services so that you have a better experience on X, including by showing you more relevant content and ads, suggesting people and topics to follow, enabling and helping you discover affiliates, third-party apps, and services. We may use the information we collect and publicly available information to help train our machine learning or artificial intelligence models for the purposes outlined in this policy. We may use the information we collect from accounts of other services that you choose to connect your X account to provide you features like cross-posting or cross-service authentication, and to operate our services. We use your contact information to help others find your account if your settings permit, including through third-party services and client applications. We use your information to provide our advertising and sponsored content services subject to your settings, which helps make ads on X more relevant to you. We also use this information to measure the effectiveness of ads and to help recognize your devices to serve you ads on and off of X. Some of our ad partners also enable us to collect similar information directly from their website or app by integrating our advertising technology.

2024-11-16

1 previously-quoted snippet(s) (wider-window capture) not present in this scrape — narrower window still captured the same core clause above.

2025-04-17

Unchanged from 2024-11-16 (see above).

2025-05-09

Unchanged from 2025-04-17 (see above).

2025-05-21

Unchanged from 2025-05-09 (see above).

2025-06-28

Clause: artificial intelligence, machine learning, train AI/models (wider-window duplicate reappears this scrape — same underlying paragraph as above, now using the updated x-services-and-corporate-affiliates link)

We think it’s most useful to describe the five main ways we use information and if you have questions that are not answered, you can always contact us. Here we go: 2.1 Operate, improve, and personalize our services. We use the information we collect to provide and operate X products and services. We also use the information we collect to improve and personalize our products and services so that you have a better experience on X, including by showing you more relevant content and ads, suggesting people and topics to follow, enabling and helping you discover affiliates, third-party apps, and services. We may use the information we collect and publicly available information to help train our machine learning or artificial intelligence models for the purposes outlined in this policy. We may use the information we collect from accounts of other services that you choose to connect to your X account to provide you features like cross-posting or cross-service authentication, and to operate our services. We use your contact information to help others find your account if your settings permit, including through third-party services and client applications. We use your information to provide our advertising and sponsored content services subject to your settings, which helps make ads on X more relevant to you. We also use this information to measure the effectiveness of ads and to help recognize your devices to serve you ads on and off of X. Some of our ad partners also enable us to collect similar information directly from their website or app by integrating our advertising technology.

2025-07-29

1 previously-quoted snippet(s) (wider-window capture) not present in this scrape — narrower window still captured the same core clause above.

2025-10-23

Unchanged from 2025-07-29 (see above).

2025-12-17

Wider-window duplicate reappears this scrape, byte-identical to the 2025-06-28 capture (see above) — not re-quoted. Core and third-party clauses unchanged from baseline.

2026-01-16

1 previously-quoted snippet(s) (wider-window capture) not present in this scrape — narrower window still captured the same core clause above.