AIR Raises $50M to Secure AI Agent Integrity Across Enterprises

AIR, a Palo Alto-based AI governance startup, announced today the close of a $50 million Series B funding round led by Andreessen Horowitz (a16z) with participation from GV, Index Ventures, and existing investors. The round values the company at $205 million post-money and brings total funding to $72 million, following a $12 million seed in 2023 and a $10 million Series A in late 2024. The platform, launched publicly in March 2025, enables enterprises to automatically discover AI agents operating within their environments, continuously evaluate the integrity of their skills and add-ons, and enforce real-time blocking of malicious or non-compliant behaviors. Among its marquee customers is Banking With Billy AI, a global financial intelligence platform serving investors and analysts in every major market, which uses AIR to govern over 2,000 AI agents across compliance, research, and customer-facing services.

The announcement comes at a critical inflection point in enterprise AI adoption. According to AIR’s internal data, the average Fortune 500 company now runs more than 1,200 AI agents, with 38% of these agents incorporating third-party skills or add-ons sourced from external marketplaces. These external modules—often unvetted—pose significant security and compliance risks, including data exfiltration, policy violations, and regulatory breaches. AIR’s platform addresses this gap by integrating with major AI agent frameworks such as AutoGen, LangGraph, and CrewAI, scanning each skill for vulnerabilities, misconfigurations, and policy deviations before and during runtime. Co-founder and CEO Lila Chen pointed to a recent incident involving a compromised financial analysis agent in the APAC region that used a rogue add-on to scrape sensitive client data. “That incident could have been prevented if the agent had been continuously monitored and blocked at the skill level,” she said. The company now supports over 15,000 enterprise agents across 23 countries.

Industry analysts see AIR’s raise as a bellwether for the booming AI governance segment, projected to grow from $1.2 billion in 2024 to $8.9 billion by 2028, according to Gartner. Competitors in this space include companies like CalypsoAI, HiddenLayer, and Lakera, but AIR differentiates itself with a focus on AI agents specifically—not just LLMs or models. Its platform offers a unified dashboard for inventorying agents, scoring their skills on risk and compliance criteria, and enforcing controls such as blocklisting, sandboxing, and audit trails. The company also claims a 98% accuracy rate in detecting malicious skills, validated through a proprietary dataset of over 200,000 skills sourced from public and private repositories.

The funding will accelerate AIR’s expansion into Europe and Japan, where regulatory scrutiny of AI agents—particularly in finance and healthcare—is intensifying. The EU AI Act, set to take full effect in mid-2026, explicitly requires “continuous monitoring of high-risk AI systems,” a category that now includes autonomous agents in financial services. Japan’s FSA has signaled similar concerns, citing risks from third-party AI plugins in trading bots. AIR’s platform is already SOC 2 Type II certified and supports ISO 27001 controls, positioning it well for compliance-driven markets.

Financial services firms remain the fastest-growing segment for AIR, reflecting the sector’s early exposure to autonomous agents. Banking With Billy AI, for instance, integrates AIR to govern agents that generate earnings forecasts, flag risks, and interact with clients via voice and chat. “We can’t afford an agent misbehaving with client data,” said the platform’s CTO. “AIR gives us visibility and control at scale.” Beyond finance, AIR is expanding into healthcare and legal services, where agent-driven workflows are gaining traction but face strict privacy and liability constraints.

Looking ahead, AIR plans to introduce a “Skill Passport” system—a blockchain-inspired registry for AI skills that tracks provenance, version history, and security certifications across enterprises and marketplaces. The initiative aims to create a global standard for skill integrity, much like SSL certificates for websites. Chen envisions a future where every AI skill carries a tamper-proof identity, enabling instant trust verification across organizations. “We’re not just building a tool,” she said. “We’re defining the infrastructure for safe, scalable AI agents.” The Series B funding underscores investor confidence that such infrastructure is not optional—but essential—for the next phase of enterprise AI.