Researchers used Claude to hack OpenAI

2026-09-20T08:51:38Z•155a657821831d323aea48aa10d2bb7d6dc85bf722e2dd6e317860d940927f9d
AI-securityASCII-smugglingBGP-hijackingChromeClickFixIoT-securityLLMMicrosoftWindowsactive-exploitationcloud-securitycybersecuritydata-breachdata-exfiltrationexploit-kitidentity-theftmacOSphishingprompt-injectionproxywaresupply-chain-attacksurveillancevulnerability

What happened

Ars Technica security feed covering active exploitation, exploit kits, large-scale breaches, supply-chain attacks, phishing and social-engineering campaigns, AI/LLM prompt-injection and data-exfiltration risks, privacy-invasive tracking, and insecure consumer devices. Multiple reports describe high-impact incidents, including a Mac vulnerability under active exploitation, attacks against corporate and public-sector infrastructure, and mass personal-data exposure.

Why it matters

A reviewed impact interpretation has not been published for this record.

Evidence and limitations

Source ID
arstechnica_security
Record identifier
155a657821831d323aea48aa10d2bb7d6dc85bf722e2dd6e317860d940927f9d
Enrichment time
2026-09-20T08:51:38Z
AI-assisted enrichment
Yes

This record may overlap with other records. Its enrichment can be incomplete or wrong, and machine assistance was used. Validate consequential decisions against the linked source and your own environment.

Record · Researchers used Claude to hack OpenAI · Baitaphish