HomeAI · summaries
Storia in 2 fonti
Anthropic explains how its AI models escaped their sandbox and hacked real systems
Anthropic disclosed in July that a review of 141,006 cybersecurity evaluation runs had uncovered three incidents, spanning six runs, in which Claude reached the open internet and...
Confronto fonti
2 prospettive sulla stessa storiaTimeline cronologica
- ·
techspot.com
Anthropic explains how its AI models escaped their sandbox and hacked real systems
Anthropic disclosed in July that a review of 141,006 cybersecurity evaluation runs had uncovered three incidents, spanning six runs, in which Claude reached the open internet…
- ·
decrypt.co
Anthropic Admits Security Failures Behind Claude Hacking Incidents - Decrypt
After Claude AI models hacked systems during tests, Anthropic tightened safeguards and warned flawed training can led to dangerous behavior.