Wallarm has launched its AI Control Platform, now available through AWS Marketplace.
The platform is designed to help organisations discover, monitor and enforce policies across AI agents, models, APIs and MCP infrastructure from a single runtime control layer. It combines AI security and API security in one system and launches with two products: AI Hypervisor and Infrastructure Discovery.
Businesses are rolling out AI tools at a pace that often exceeds their ability to track what is in use or how those systems behave in production. Wallarm cited figures showing that nearly 80% of organisations have experienced data incidents involving generative AI, while 72% of corporate AI tools in active use are rated high or critical risk.
Wallarm also pointed to findings from AWS's 2025 Generative AI Adoption Index, which found that 45% of organisations now rank generative AI as a top IT budget priority. That mix of spending and risk has increased demand for tools that monitor AI systems as they run rather than through periodic checks.
The company says the platform is intended to close that gap by identifying AI infrastructure across AWS environments in real time, tracing AI sessions end to end and applying policy controls when risks are detected. It is aimed at helping security and IT teams answer practical questions about which assets make up their AI stack, what data those systems access and whether sensitive information is leaving the organisation.
Built on AWS
AWS provides the underlying infrastructure and native security services, while Wallarm's software extends visibility into AI behaviour, including the data AI systems access and the external services they call. The setup is intended to give customers a path from discovery to runtime governance and enforcement without changing application code.
One of the first components, AI Hypervisor, is designed to discover AI agents, frameworks, model provider calls, MCP servers, APIs and data sources based on runtime behaviour. It monitors AI workloads on Kubernetes and records outbound connections to model providers including AWS Bedrock, Anthropic, OpenAI and Azure OpenAI.
AI Hypervisor traces each AI session from the initial prompt through model and tool calls back to the user. It also detects sensitive data in transit, attributes AI usage and cost to the responsible user or team, and can enforce policy at the connection level.
The second launch product, Infrastructure Discovery, focuses on mapping AWS environments. It continuously charts AWS accounts, regions and resources across compute, networking, API Gateway, Lambda and IAM, then builds a relationship graph showing how systems connect.
It also detects changes between scans and places AWS Security Hub findings directly onto the relevant assets, with descriptions rewritten in plain language for security teams. The goal is to help teams identify which issues need immediate attention across a large cloud estate.
The broader backdrop is growing regulatory pressure. Wallarm says organisations in regulated sectors face a near-term compliance deadline under the EU AI Act, which will require companies to demonstrate AI governance and oversight.
"AI adoption is outpacing governance, and customers are being forced to trade speed for control. The AI Control Platform removes that tradeoff for every CIO scaling AI and every CISO governing it. This is the foundation Wallarm is building on, and the capabilities we deliver across this year and next will compress the time between AI deployment and AI governance to near zero," said Shayne Higdon, Chief Executive Officer, Wallarm.
Wallarm describes the platform as the foundation of a broader product line that will expand over time. For now, it is positioning the offering around continuous oversight of AI systems in production, with a focus on visibility, policy enforcement and audit evidence.
One customer cited by Wallarm linked the platform to the challenge of securing a more complex technology environment as AI adoption spreads across production systems and interfaces.
"As CarCollect scales, Wallarm helps us secure our growing ecosystem of AI, MCP servers and APIs from endpoint risks and potential sensitive data exposure. Wallarm gives us the full visibility and control we need to confidently strengthen our overall security posture," said Wout Groenendijk, Chief Technology Officer, CarCollect.