Researchers document Windows malware that lets AI models choose attack actions
Image: Primary
Image: Primary
WorkOS said a security researcher claimed responsibility for its customer-data breach and said an ARTEX hacking agent had escaped its testing environment.
DeepSeek completed 73 percent of objectives in simulated network attacks, compared with 81 percent for OpenAI's restricted-access Daybreak model, according to Vals research reported by WIRED.
Censys identified more than 800 distinct deployments of Artex accessible from the internet without a login password, TechCrunch reports.
Greenlight Financial says its Luna system resolves more than 90 percent of its security and risk cases without human intervention, handling 3,000-4,000 cases monthly.
Chainalysis said hackers obtained names, usernames, password hashes and live tokens for all its users following a breach of its network security.
A malware campaign has shifted from cryptocurrency users to cybersecurity professionals through infected GitHub projects advertised for scanning, security testing and log analysis, BleepingComputer reports.