Joe Sullivan joins Manifold Security board as AI agent risks grow

Manifold Security appointed former Facebook, Uber and Cloudflare CSO Joe Sullivan to its board as the startup expands tools for securing AI agents across endpoints, browsers, desktop apps and cloud environments. The move comes as enterprises rush to deploy agents and security teams face new demands for visibility, governance and real-time response.

Why it matters: - AI agents are moving from experiments into enterprise workflows, creating new security risks across employee endpoints, browsers, desktop applications and cloud environments. - Manifold Security is positioning itself as a control layer for that new workload, and Sullivan brings board-level credibility from some of the largest security programs in tech. - The appointment signals that agentic AI security is becoming a category where enterprise customers and investors expect proven operators, not just startup founders.

What happened: - Manifold Security named Joe Sullivan to its Board of Directors on Sept. 10, 2026. - Sullivan previously served as CSO at Facebook, Uber and Cloudflare. - Sullivan also worked as a federal prosecutor at the U.S. Department of Justice and helped establish the Computer Hacking and Intellectual Property Unit in the Northern District of California. - Manifold said Sullivan is an early backer of the company and one of the investors behind the $15 million raised to date.

The details: - Manifold Security describes itself as an agentic security company focused on protecting organizations from risks created by AI agents. - The company is scaling a platform for securing AI agents operating across employee endpoints, browsers, desktop applications and cloud environments. - Manifold said it has already secured the agentic workforce of multiple Fortune 500 companies. - The company is extending coverage wherever agentic work takes place inside the enterprise, including on the endpoint, in the browser and in the cloud. - Deloitte has projected that 74% of companies expect to use AI agents by 2027. - Manifold said enterprises need three core capabilities: full visibility into agents, governance over agent behavior and real-time threat detection and response. - The company’s leadership team includes CEO Neal Swaelens and CTO Oleksandr Yaremchuk. - Before founding Manifold, Swaelens and Yaremchuk launched Laiyer AI, where they created LLM Guard, described by the company as the most widely adopted open-source LLM firewall.

Between the lines: - Sullivan’s comments frame AI agents as probabilistic systems that can behave correctly most of the time and still create outsized risk when they go off course. - His board role suggests Manifold wants to pair technical product development with executive-level security experience as the market matures. - The company is also leaning into a broader industry narrative: security teams need visibility and control before agent adoption scales further.

What's next: - Manifold will keep expanding coverage for agentic work across the enterprise as adoption grows. - Sullivan said he is joining the board to help the team as the market develops. - The company is building what it describes as the security layer for the agentic workforce as organizations move faster on AI deployment.

The bottom line: - Manifold is betting that the next security platform category will be built around controlling AI agents, and Joe Sullivan’s board seat is meant to reinforce that pitch.