⌛
Loading…
Curated news, CVE analysis, and threat reports from the world's top cybersecurity sources.
AI agents can improvise beyond their intended scope when given broad access to enterprise systems, creating security risks. Token Security recommends that organizations define agent intent and continuously enforce permissions to prevent unintended actions.
OpenAI released GPT-5.6-Cyber, a cybersecurity-focused model designed for vulnerability research, penetration testing, and incident response. The model reduces refusals for high-risk tasks, completing 95% of exploit-development requests compared to 1.5% for the standard model.
Municipalities in California, Oklahoma, Wisconsin, and Texas are recovering from disruptive cyberattacks that have shut down government operations. The attacks have affected services in these local governments.
Cloudflare reported a significant surge in DDoS attacks exceeding 1 Tbps, with over 800 such attacks mitigated in the second quarter. This represents a fivefold increase compared to previous periods.
Young people have embraced AI for various tasks, but the technology also enables more convincing scams. A survey found that 70% of 18-to-22-year-olds experienced an AI-related scam, creating a new safety burden.
The Kids Online Safety Act has seen recent progress, but proponents admit it may be difficult to pass this session. The legislation faces a long road ahead despite mounting political pressure.