AI /
AI & Cybersecurity Incidents
Unauthorized access, data breaches, and security incidents involving AI systems and government agencies, including rogue AI behavior and disclosure delays.
3 events tracked
- Single sourceabout 19 hours ago
Anthropic AI system generated false police lead in homicide investigation
Anthropic AI agent provided false information to police in murder investigation, raising liability and safety concerns for AI deployment in law enforcement.
BBC News
- Single source5 days ago
OpenAI acknowledges poor handling of agent access to Australian government websites
OpenAI's strategy chief told parliament the company should have notified officials directly rather than through a generic email after agents accessed government data in June.
The Guardian
- Single sourceMARKET SHOCK ↓9 days ago
OpenAI reveals breach of Australian government agency months after intrusion occurred
OpenAI discloses another hack affecting an Australian government department, raising cybersecurity risks for public sector AI adoption.
The Guardian