Mate Security has raised $35 million in Series A funding as enterprises look for new ways to operate security teams against attacks moving at machine speed.
Canaan Partners led the round, with participation from Insight Partners, Team8 and M12, Microsoft’s (MSFT ) venture fund. All of Mate’s existing investors returned for the financing, bringing the cybersecurity company’s total funding to more than $50 million.
The Series A arrives only eight months after Mate Security emerged from stealth with an oversubscribed $15.5 million seed round. The company says it has grown by more than 500% since the third quarter of 2025, driven in part by adoption among Fortune 500 security operations centres.
Mate Security plans to use the capital to meet growing enterprise demand, expand into additional markets and continue developing its agentic security operations platform.
Security Operations a New Scaling Problem
The rapid development of generative AI is affecting both sides of cybersecurity.
Security teams are gaining access to AI systems capable of analysing alerts, collecting evidence and automating portions of an investigation. At the same time, attackers can use AI to probe systems, adapt techniques and launch campaigns at a speed that traditional security operations were not designed to handle.
This creates a difficult mismatch. Many security operations centres, or SOCs, still rely on analysts moving between disconnected tools, reviewing queues of alerts and manually reconstructing what happened during an incident.
Adding an AI assistant to that environment does not necessarily solve the underlying problem. An agent may be able to summarise an alert or query a security platform, but it can still reach the wrong conclusion when it does not understand how the organisation actually operates.
Mate Seucrity is trying to address that limitation by giving security agents access to business and operational context rather than asking them to reason from isolated security events.
“AI is forcing a fundamental rethink of security operations,” said Joydeep Bhattacharyya, general partner at Canaan. “What stood out to us about Mate wasn’t simply its use of AI; it was the team’s conviction that trustworthy AI requires a deep understanding of how an organisation operates.”
A Security Context Graph Sits at the Centre of the Platform
Mate Security’s platform is built around what the company calls a Security Context Graph.
The graph collects and organises knowledge from across an organisation, including information about users, assets, policies, previous investigations, operational workflows and approved activities. Mate Security describes these pieces of information as memories that can be connected and queried by its AI agents.
Rather than treating every login, file transfer or configuration change as an isolated event, the system attempts to determine how the activity fits into the wider business.
A series of suspicious login attempts, for example, could appear malicious when viewed only through identity logs. The same activity may be less concerning when the system knows that an authorised penetration test was scheduled for that period.
Similarly, an employee downloading sensitive files could indicate data theft. It could also be consistent with an approved project or change in responsibilities. Mate Security’s agents are designed to consider factors such as personnel changes, document classifications and earlier investigations before reaching a verdict.
The company says the graph is continuously rebuilt as policies, asset ownership and operating conditions change. Completed investigations can also feed new information back into the platform, allowing later investigations to draw on the organisation’s accumulated security knowledge.
Mate Security argues that this data-first architecture can make AI-generated verdicts more accurate, consistent and explainable. It also helps preserve institutional knowledge that might otherwise disappear when experienced analysts leave an organisation.
Specialised Agents Cover the Investigation Cycle
On top of the context layer, Mate Security operates specialised agents for detection engineering, alert triage, investigation, threat hunting and incident response.
The platform is intended to connect those functions as a continuous process rather than treating each as a separate workflow. An agent investigating an alert can query evidence from the organisation’s existing technology stack, use the Security Context Graph to interpret what it finds and pass structured information to another agent responsible for the next stage.
Mate Security says its agents can work with security information and event management platforms, endpoint detection tools, identity systems, ticketing platforms, communications applications and standard operating procedures without requiring every piece of data to be moved into a single repository.
Response actions remain governed by permissions. Depending on how the organisation configures the system, Mate Security can recommend a containment step, request human approval or execute an authorised action.
This reflects what the company calls a “least-agency” principle. Each agent receives only the permissions and contextual information required for its assigned task, reducing the risk created by giving a broadly autonomous system unrestricted access.
The company’s broader goal is to let security teams gradually expand automation as the platform demonstrates that it can make reliable decisions. High-impact actions can remain supervised, while routine investigations may be handled with less analyst involvement.
An Open Approach to Agentic Security
Mate Security is positioning the platform as an open foundation rather than a closed collection of proprietary agents.
Customers can use Mate Security’s own agents, connect specialised agents from other security vendors or build agents internally. Each can operate on the same governed context layer and under common controls for permissions, quality, auditability and response.
That architecture addresses an emerging concern as enterprises adopt more AI tools: agent fragmentation.
Without a shared layer, individual security agents may gather their own data, maintain separate memories and apply different rules when making decisions. This can produce conflicting conclusions and make it difficult for security leaders to understand why an automated action was taken.
Mate Security’s orchestrator is intended to provide a common operating structure so that agents use the same organisational knowledge and trust controls, even when they were developed by different providers.
The company also states that customers retain ownership of their context and are not locked into a single agent, AI model or security vendor. Its founders describe that openness as a prerequisite for organisations that want to take advantage of future advances in AI without rebuilding their security architecture each time a new model emerges.
AI Agents Could Reshape Security Operations
Platforms such as Mate Security point towards security operations becoming more continuous, contextual and automated. Instead of investigating alerts in isolation, AI agents could retain lessons from previous incidents and use them to improve future detection, triage and response.
This could shift analysts away from repetitive evidence gathering and towards supervising agents, investigating unfamiliar threats and deciding when automation should be allowed to act.
A shared context layer may also become important as enterprises deploy agents from multiple vendors. Without a common understanding of users, assets and policies, those agents could duplicate work or reach conflicting conclusions.
The likely outcome is not a fully autonomous security operations centre, but a hybrid model in which agents handle routine investigations while humans retain control over ambiguous or high-impact decisions.
