General Analysis, a San Francisco-based cybersecurity startup focused on securing agentic artificial intelligence systems, has emerged from stealth with $10 million in seed funding to accelerate development of its AI security platform. The round was led by Altos Ventures, with participation from 645 Ventures, Menlo Ventures, Y Combinator, and additional strategic investors and angel backers. The company said the funding will be used to expand its engineering team, deepen product development, and grow customer deployments as enterprises increasingly adopt autonomous AI agents.

Founded by former researchers and engineers from NVIDIA, Cohere, and DeepMind, General Analysis is building security infrastructure designed specifically for AI agents capable of taking autonomous actions on behalf of users and organizations. As enterprises deploy AI systems that can access sensitive data, execute workflows, and interact directly with customers, the company believes traditional cybersecurity tools are insufficient to address the new risks introduced by agentic AI. Its platform is designed to identify vulnerabilities before AI agents are deployed into production environments.

The startup’s technology combines automated red teaming, runtime guardrails, interpretability tools, and observability capabilities to evaluate how AI agents behave under real-world conditions. Rather than focusing exclusively on conventional software vulnerabilities, General Analysis examines how large language model-based agents respond to adversarial prompts, malicious users, and unexpected situations that could lead to financial losses, data exposure, or unintended actions. The platform enables organizations to identify weaknesses before they become operational security incidents.

General Analysis has demonstrated the need for this type of testing through research highlighting the vulnerabilities of production AI systems. In one widely cited experiment, the company’s adversarial AI agent successfully persuaded 50 live customer service bots to provide more than $10 million in unauthorized benefits, including high-value gift cards and extended free services. According to the company, only five of the 55 systems tested resisted the simulated attacks, illustrating the growing gap between the rapid adoption of AI agents and the security controls available to protect them.

The company’s founders argue that AI agents introduce fundamentally different security challenges from traditional software because they make probabilistic decisions, interact with natural language, and continuously adapt to changing inputs. Existing security testing methods were largely developed for deterministic applications, making them less effective when evaluating autonomous AI systems capable of reasoning and executing multi-step tasks. General Analysis aims to provide enterprises with tools that continuously evaluate AI behavior before and after deployment, helping organizations maintain confidence in production AI environments.

The newly raised capital will support continued product development as the company expands its customer base across industries including customer support and financial services. General Analysis said its platform is already being used by enterprise customers whose products and workflows serve hundreds of millions of users. These organizations are deploying AI agents to automate increasingly complex business operations while seeking assurances that the systems remain secure, reliable, and resistant to adversarial attacks.

Investor interest in General Analysis reflects broader market demand for AI-native cybersecurity solutions. As enterprises move beyond experimental AI deployments toward autonomous systems capable of performing business-critical functions, security has emerged as one of the industry’s highest priorities. Venture firms backing the company believe organizations will require dedicated infrastructure to validate AI behavior before granting autonomous systems access to sensitive business processes.

The funding also comes as regulators and enterprise customers place greater emphasis on AI governance, transparency, and risk management. Organizations adopting agentic AI increasingly face pressure to demonstrate that autonomous systems operate safely while complying with internal policies and external regulatory requirements. General Analysis believes continuous testing, monitoring, and adversarial evaluation will become essential components of enterprise AI deployment, much like penetration testing and vulnerability assessments are for conventional software.

Chief Executive Officer Rez Havaei said the company intends to build the security layer that enables organizations to deploy AI agents with confidence rather than hesitation. By identifying vulnerabilities before attackers exploit them, General Analysis hopes to reduce operational risks while accelerating responsible AI adoption across enterprises. The company expects the new funding to support continued expansion of its engineering organization, customer relationships, and research into securing increasingly capable autonomous AI systems. With backing from Altos Ventures and a syndicate of experienced early-stage investors, General Analysis is positioning itself to become a foundational provider of security infrastructure for the rapidly evolving agentic AI ecosystem.

Related Articles