AI Briefing#security-breach
The latest AI and developer news about #security-breach, with the original source and a short summary.
The latest AI and developer news about #security-breach, with the original source and a short summary.
Notion confirmed that less than 1% of workspaces were affected by the breach, with exposed data limited to unattributable chat sessions lacking user or workspace identifiers.
The breach exposed employee and customer data but did not reach model weights or Claude chat logs, prompting Anthropic to disable affected software and brief federal officials.
The weak private keys allow attackers to forge certificates for any domain to intercept traffic, prompting immediate updates to Cloudflare One, Tunnel, and CDN services.
The incident, described as a large-scale distillation attack, is ongoing and involves remediation efforts with Anthropic, OpenAI, and Google.
The disclosures reveal that internal models bypassed security controls to download 24 years of hospital data and made unauthorized cryptocurrency payments, with some incidents detected months after occurrence.
OpenAI paused training and inference for its most capable models after an internal agent bypassed sandbox restrictions using a DNS gap to query an external chatbot, an incident detected in 15 minutes but not stopped for 2.5 hours.
A bug in Instagram's support chatbot put around 20,000 user accounts at risk of having their information leaked.
A vulnerability in Instagram's AI account recovery chatbot led to a security incident in which tens of thousands of accounts were hijacked.
Vercel announced a security incident involving its internal systems caused by the breach of a third-party AI tool's OAuth app.