Straiker, a Sunnyvale, California-based provider of an AI-native security platform for agentic applications, has raised $64 million in Series A funding. The platform secures coding agents like Claude Code and Cursor, productivity agents such as ChatGPT and Copilot, and custom agents built on LangChain or AWS Bedrock against threats including prompt injection, tool poisoning, and memory tampering. The capital will expand product development across its Discover AI, Ascend AI, and Defend AI offerings.
Agentic AI Security Category Emerges
The timing aligns with rapid agent deployment and high-profile incidents. Cranium raised $46 million total including a $25 million Series A in October 2023. Prompt Security secured $23 million before its acquisition by SentinelOne, while Lakera raised $30 million prior to its Check Point acquisition. Straiker's full-lifecycle approach addresses gaps in visibility, pre-deployment testing, and runtime protection that narrower tools leave exposed.
Unique Attack Vectors Target Agents
Straiker's research shows 28.6 percent of cataloged MCP tools are dangerous by design. Additionally, 91 percent of attacks on productivity agents result in silent data exfiltration, and 36 percent of successful attacks on coding agents lead to remote code execution. Existing security solutions built for static applications fail to handle non-deterministic agent behavior or inter-agent manipulation.
Three-Product Stack Covers Lifecycle
Straiker built Discover AI for agent discovery and governance, Ascend AI for adversarial red-teaming using fine-tuned offensive models, and Defend AI for sub-second runtime guardrails. The company uses a medley-of-experts architecture rather than relying on frontier LLMs, achieving 98.1 percent detection accuracy in benchmarks. This contrasts with incumbents retrofitting existing platforms and acquired startups that lost independent focus after consolidation.
"We are entering an era where I believe 80 percent of the organization will be run on agents. Humans will be behind them," said CEO Ankur Shah.
Tier-1 Investors Signal Platform Potential
Marathon Management Partners led the round with participation from Citi Ventures, Illuminate Financial, Workday Ventures, and continued backing from Bain Capital Ventures and Lightspeed. Gokul Rajaram joined the board. The syndicate reflects conviction in infrastructure plays that scale across enterprise agent ecosystems.
Market Expands Rapidly Amid Consolidation
The agentic AI security market is projected to grow from $1.65 billion in 2026 to $13.52 billion by 2032 at a 42 percent CAGR. The broader agentic AI market stands at $9.14 billion in 2026 and is expected to reach $139.19 billion by 2034. With over $3.6 billion in startup funding across the space and multiple acquisitions, capital is flowing toward specialized platforms that handle the full agent attack surface.
Research-Driven Differentiation
Co-founders Ankur Shah, formerly SVP and GM of Prisma Cloud at Palo Alto Networks, and Sreenath Kurupati, ex-VP of AI and Security Research at Akamai and founder of Cyberfend, bring deep domain expertise. Their STAR Labs team publishes original vulnerability research that informs product development and establishes category leadership.