Choose a continent to filter news, research and evidence analysis.
Latest verified release
7 Nov, 16:00 GMT
Hourly editorial reviews are scheduled. Reports appear after verification; source dates and evidence limits stay visible.
Read in your language
Choose one of 20 languages. Translations open in Google Translate and may contain errors. Source links remain available.
Only the public page is shared with Google when you choose a language. Your browser language suggests a choice; travel never changes it automatically.
Viewing Global
Dangerous AI, without the panic
Follow verified reporting on cyber misuse, fraud, deepfakes, unsafe autonomy and frontier-model safety. We separate observed harm from plausible risk and speculation, with the original evidence attached.
Global intelligence feed
34 recent verified reports
Reports appear after verification. Current coverage uses sources from today and the preceding three calendar days. Earlier work is clearly dated. Browse the archive.
Technology · Announced
Phone, tablet and wearable users may gain clearer compatibility and app-access boundaries, but the notice does not show how understandable disclosures will be.
1 min brief · 8 min full report · source 2026-11-06
Security & Defence · Announced
People affected by a data leak may gain clearer routes to police, consumer and specialist support, but the package has not yet shown how quickly help will arrive or whether secondary fraud will fall.
2 min brief · 10 min full report · source 2026-11-06
Security & Defence · Studied
Developers using third-party LoRA adapters should treat them as executable supply-chain components and preserve provenance, hashes and isolated test results.
2 min brief · 10 min full report · source 2026-11-06
Technology · Studied
Developers and small organisations may gain access or simpler billing through a relay, but can face hidden context, unstable behaviour or effective costs that the advertised unit price does not reveal.
2 min brief · 11 min full report · source 2026-11-05
Security & Defence · Announced
Security and audit staff gain a common structure for evidence, but will need access to models, logs, suppliers and sector expertise to apply it credibly.
2 min brief · 11 min full report · source 2026-11-06
AI Risks & Safety · Announced
Workers overseeing agent teams need action-level logs and clear escalation routes; chat summaries alone may hide both sabotage and ordinary mistakes.
1 min brief · 9 min full report · source 2026-11-06
Useful context with its original source date. These reports are outside the current news window.
Original source 2026-11-03
Original source 2026-11-03
Original source 2026-11-02
Google’s consent message lets readers in the UK, EEA and Switzerland allow, decline or manage advertising consent. You can reopen it to change or withdraw consent.