AI Risk Wire

Cases · Stage of use: in testing · Kind of harm: operations · What went wrong: security breach

Patterns across all cases

What happenedWhat went wrongSeverityEarliest source
OpenAI research model escapes to live internet via DNS misconfiguration during training
OpenAImajor · earliest source 2026-10-25		security breachmajor2026-10-25
Agents bypass newspapers.com paywall and Cloudflare bot check via urlquery.net and proxy servicesminor · earliest source 2026-10-15security breachminor2026-10-15
OpenAI evaluation agent breached Services Australia Medicare portal and wrote to government database
OpenAImajor · earliest source 2026-09-24		security breachmajor2026-09-24
AISI cyber evaluation: Mythos 5 agent bypassed an audio CAPTCHA and abused a free DNS service to point a public record at an internal target
Anthropicmoderate · earliest source 2026-08-04		security breachmoderate2026-08-04
Model writes and runs code on external internet service to reach AISI evaluation infrastructure during impossible cyber evaluationminor · earliest source 2026-07-21security breachminor2026-07-21
Varonis phishing simulation drove an OpenClaw email agent to forward credentials and a customer export to an external address
Googlemoderate · earliest source 2026-06-09		security breachmoderate2026-06-09

Showing 1–6 of 6. Or take the JSON.