What happened
June 29, 2026: Straiker announced a $64M Series A (total $85M), led by Marathon Management Partners with Bain Capital Ventures and Lightspeed continuing. The GA platform comprises three modules — Discover AI (agent inventory and MCP visibility), Ascend AI (pre-deployment adversarial red-teaming of agent tools, MCP servers, and workflows), and Defend AI (runtime blocking of prompt injection, data leakage, and tool manipulation). STAR Labs research found 91% of attacks on productivity agents resulted in silent data exfiltration and 36% of coding-agent attacks achieved RCE. Run-rate revenue grew 15× in under a year; customers include Fortune 500 and frontier AI labs.
Why it matters
Largest dedicated agentic-security funding round to date, from a team with Prisma Cloud and Akamai security pedigree. The three-layer Discover/Red-team/Runtime architecture is the most complete purpose-built stack for agentic AI security currently GA, filling the gap left by general-purpose tools not designed for agent autonomy.
Applicability
Enterprises deploying coding agents (Claude Code, Cursor) or productivity agents (ChatGPT, Copilot) at scale should evaluate Straiker for pre-production red-teaming and runtime guardrails; especially urgent for orgs without existing agent security controls.