API Attacks Surge as AI Expands the Blast Radius; Wallarm Flags MCP Risk
APIs were the leading exploitation vector in 2025, with Wallarm finding ~11,000 API-related flaws from 60,000 disclosures and CISA data linking APIs to 43% of actively exploited cases. Advances in generative AI and coordinating agents are compressing the time from disclosure to weaponized exploit and amplifying social-engineering value, pushing defenders toward runtime enforcement, behavioral telemetry, and identity-first controls.
