Sweet Security's AI 'Attacker' Fortifies Cloud Defenses

Sweet Security has launched Sweet Attack, an AI-powered agent designed to simulate an adversary within a company's live systems. This new tool helps fortify cloud defenses against escalating AI-driven attacks. Sweet Attack aims to eliminate 90% of the time attackers spend guessing. It can surface exploitable attack paths in just three days that were previously unidentified. This comes after Anthropic's Mythos model autonomously found and exploited software vulnerabilities, including a 17-year-old bug in FreeBSD. Experts say Sweet Attack marks a critical shift in cybersecurity. It allows organizations to proactively validate and prioritize real threats. This matters because it helps businesses prepare for a new class of AI threats.