Ambient.ai's AI Guardians Signal New Era for Physical Security

REDWOOD CITY, Calif. – February 18, 2026 – A fundamental shift is underway in how global enterprises protect their people and assets. Ambient.ai, a pioneer in artificial intelligence for physical security, today announced financial results that signal a decisive market turn away from reactive monitoring toward proactive, AI-driven threat prevention. The company has doubled its new annual recurring revenue (ARR) in the last fiscal year and boasts a net revenue retention rate exceeding 140%, indicating that its largest customers are rapidly expanding their use of the platform.

This explosive growth, which includes a tripling of its customer base, is built on the promise of what the company calls 'Agentic Physical Security.' This new paradigm leverages advanced AI that can perceive, reason, and act autonomously, transforming standard security cameras from passive recorders into intelligent, ever-vigilant guardians. The results suggest that the inflection point has been reached, with major corporations now betting big on AI to not just see what happens, but to understand what is happening and prevent incidents before they occur.

The Rise of the Reasoning AI

At the heart of Ambient.ai's recent success is a significant technological leap forward with the introduction of Ambient Pulsar. The company describes it as the first edge-optimized, frontier reasoning Vision-Language Model (VLM) purpose-built for the complexities of physical security. Unlike general-purpose AI models, Pulsar has been trained on an immense, specialized dataset of over one million hours of real-world enterprise video.

This specialized training allows Pulsar to achieve a level of contextual understanding that has long been the domain of human operators. Instead of just flagging motion, the system is designed to reason about a scenario. It can differentiate between a janitor propping a door open for a cleaning cart and an unauthorized attempt at tailgating, or identify the subtle signs of a potential safety hazard before it escalates into an accident. Ambient.ai claims that in physical security use cases, Pulsar's reasoning performance exceeds that of even highly anticipated general models like OpenAI's GPT-5 and Google's Gemini 2.5 Pro, while operating at up to 50 times greater efficiency—a critical factor for real-time processing across thousands of cameras.

"Organizations recognize that reactive security is no longer sufficient and look for AI that can truly understand context, reason through complex scenarios, and take autonomous action to prevent incidents before they escalate," said Shikhar Shrestha, CEO and Co-Founder of Ambient.ai. "That shift is happening now and Ambient.ai is leading the way."

From Cost Center to Strategic Asset

The market's validation of this approach is evident in the numbers. Ambient.ai's financial momentum, achieved with a lean sub-1x burn multiple rare for a venture-backed startup, points to a sustainable and highly efficient business model. More importantly, it reflects the tangible return on investment its customers are experiencing. The platform is now trusted by a growing roster of industry leaders, including Cisco, ServiceNow, SentinelOne, and TikTok, as well as public sector entities like Norwalk Public Schools.

One of the most compelling examples comes from ServiceNow's Global Security Operations Center (GSOC). By deploying Ambient.ai, the enterprise software giant achieved a staggering 94% reduction in false alarms. This single metric translated into saving 15,069 hours of manual triage, freeing up highly trained security personnel from the monotonous task of sifting through endless, meaningless alerts. This shift doesn't just save money; it transforms the GSOC from a reactive call center into a proactive strategic hub focused on protecting people and assets.

Previous data from 2022 further supports this economic argument, showing that customers realized over $90 million in projected operational cost savings. This ROI is derived from preventing security breaches, optimizing guard deployments, and streamlining investigations, fundamentally changing the economics of physical security from a necessary cost center into a source of operational efficiency and measurable value.

A New Human-AI Partnership

Contrary to fears of AI replacing human jobs, the agentic security model fosters a powerful human-AI partnership. The technology acts as an intelligent assistant, augmenting the capabilities of human security teams rather than supplanting them. By autonomously identifying and clearing millions of non-critical alerts, the platform eliminates the 'alarm fatigue' that plagues traditional security operations and can lead to missed critical events.

When a high-severity incident is detected, the AI doesn't just sound a generic alarm. It provides security operators with immediate, context-rich information, enabling a response that is reportedly up to 10 times faster. This allows human teams to do what they do best: make critical judgments and intervene effectively in high-stakes situations. The role of the security professional evolves from that of a passive monitor staring at screens to an empowered, data-informed responder.

This collaborative approach is a key part of navigating the complex and high-stakes world of modern security. The market itself is undergoing a major transformation, with the global cybersecurity market—which is increasingly converging with physical security—projected to reach over $350 billion by 2030. In this environment, solutions that can provide a unified intelligence layer over existing infrastructure are highly sought after. Backed by top-tier venture firms like Andreessen Horowitz, Ambient.ai is positioned not just as a technology provider, but as a central player in defining the future of a safer, more efficient, and more intelligent physical world.