VentureBeat August 8, 2024
Anthropic, the artificial intelligence startup backed by Amazon, launched an expanded bug bounty program on Thursday, offering rewards up to $15,000 for identifying critical vulnerabilities in its AI systems. This initiative marks one of the most aggressive efforts yet by an AI company to crowdsource security testing of advanced language models.
The program targets “universal jailbreak” attacks — methods that could consistently bypass AI safety guardrails across high-risk domains like chemical, biological, radiological, and nuclear (CBRN) threats and cybersecurity. Anthropic will invite ethical hackers to probe its next-generation safety mitigation system before public deployment, aiming to preempt potential exploits that could lead to misuse of its AI models.
AI safety bounties: A new frontier in tech security
This move comes...
Today's Sponsors
Today's Sponsor