Lasso Security closed a $30 million Series A on September 2, 2026, a bet by investors that the rush to deploy autonomous AI agents has created security gaps traditional tools can't plug. ClearSky led the round, joined by Entrée Capital, iAngels, Singtel Innov8, Mindset Ventures and Swish Data, bringing the Tel Aviv-based company's total raised to $36 million since it surfaced from stealth in November 2023.
The timing is no accident. Lasso's raise lands during what looks increasingly like an arms race in agentic security, with HiddenLayer announcing a $100 million Series B on September 2, 2026, and AIR unveiling a $50 million seed on September 1, 2026. That's on top of sizable rounds from Mindgard and Geordie AI earlier in the year. The message from capital markets is unambiguous: enterprises are throwing AI agents into production, often faster than their security teams can vet them.
Alongside the funding, Lasso unveiled LEAP, a CPU-based guardrail system the startup claims can evaluate AI requests in under five milliseconds. Most competing guardrails lean on GPU-dependent large language models, which introduces latency and cost. LEAP instead routes the bulk of traffic through what Lasso calls a "transformer-free engine," escalating only edge cases to RAPID, its self-hosted LLM arbiter. The company says the architecture delivers throughput gains measured in the thousands compared to existing products, though independent benchmarks remain scarce.
"Two years ago, the question was whether enterprises would put AI in front of their customers," said Elad Schulman, Lasso's co-founder and CEO. "That question is settled. What they need now is security that keeps up with agents that act autonomously and at scale."
Schulman isn't exaggerating the scope of the problem. Lasso's platform now protects what the company describes as tens of thousands of agents handling billions of requests and actions monthly. The customer roster includes BMW, the U.S. Department of Homeland Security, Leonardo Defense, eToro, Fiverr, financial services firm Kaufman Rossin and consulting partner Protiviti. Whether that concentration in defense and financial services reflects Lasso's go-to-market strategy or a broader pattern of which sectors are willing to bet early on agentic AI remains an open question.
Vladislav Rudnitsky, CISO at Kaufman Rossin, credited Lasso with giving his team "visibility into how artificial intelligence is being used across the firm, which enables faster security decisions with empirical data." That kind of discovery capability matters more as AI proliferates beyond controlled pilots into the messy reality of departmental shadow IT.
Lasso's platform connects discovery, risk management, automated red teaming and runtime protection. It scans organizations for AI usage, tests agents for vulnerabilities, and monitors live requests to block jailbreaks, prompt injections and unauthorized data access. The company holds SOC 2 Type II certification and sells through AWS Marketplace and Microsoft's commercial marketplace, standard table stakes for enterprise security vendors.
According to Lasso, revenue grew more than 500% over the past year, though the company declined to disclose its valuation or absolute revenue figures and the claim is not independently verified. LinkedIn shows a range of 11 to 50 employees, though some third-party trackers estimate headcount in the low-to-mid 70s.
The federal push has been deliberate. Lasso formed a separate entity, Lasso Federal LLC, to serve U.S. public sector customers and partnered with Swish Data to deliver AI security to federal agencies. Public contract data shows a $1.4 million delivery order for USCIS tied to that partnership, according to USAspending records. Swish has since joined as an investor.
Ryan Crum, a ClearSky partner and former CISO of Apollo Global Management, joined Lasso's board with the round. "AI agents are going into production across every major enterprise, and most of them have never been properly tested for security," Crum said. "As a former CISO, that's a problem I've felt firsthand."
The company plans to pour Series A proceeds into engineering, deeper work in federal and regulated industries, and expanded go-to-market operations across North America and Europe. LEAP is already in production at global enterprises and within the U.S. federal government, according to Lasso, though the company did not specify which agencies beyond DHS and USCIS.
Lasso has also woven integrations with Cloudflare for network-level monitoring and CrowdStrike Falcon for endpoint telemetry. The startup's board includes former Israeli Prime Minister Naftali Bennett and Axonius co-founder Dean Sysman, a signal of the kind of political and enterprise credibility the company is assembling.
Standards bodies are catching up. OWASP released its 2026 GenAI Top 10 and a new Agent Control Standard in September, an acknowledgment that agentic attack surfaces need their own taxonomy. Whether Lasso's CPU-first architecture becomes the dominant approach or just one path among many will depend partly on how well LEAP holds up under the kind of adversarial testing that security researchers inevitably bring to new defenses.
For now, the capital flowing into this corner of the market suggests investors believe the problem is real and the window is open. Schulman said the company's next milestone centers on expanding LEAP's capabilities and widening its federal footprint, a combination that positions Lasso at the intersection of two of the fastest-moving forces in enterprise tech: the agent revolution and the U.S. government's scramble to secure it.
