How we report

What we cover

AI Safe Mode covers real incidents involving AI, especially AI agents that can browse, click, buy and run commands on their own. We focus on what happened, why it matters to ordinary people, and what you can do about it.

Sourcing and verification

• Every report is based only on published news reports. We do not publish rumours or unverified claims.
• Every claim is attributed to the outlet and date it came from, and each report cites at least two independent outlets.
• Each report is checked against its sources by a separate review step before it is published.

No names

We don't name the companies, products or people involved in incidents. The point is the lesson, not blame. Our cited sources name them where relevant.

Always a fix

Every report ends with practical steps you can take to stay safe.

Dates and updates

Each report shows the date it was published. If a report is changed after publication, it shows an updated date, and material changes are handled under our corrections policy.

Independence

Reports are not paid for or sponsored. Embedded videos and posts are our own.

Contact

Questions or errors: [email protected]. See also About & rules.