Harmonic Security Raises $50M Series B for Workforce AI

The person renewing an AI subscription could need the same evidence as the person investigating a data leak. Harmonic Security is building a business around that overlap, using its understanding of workplace AI activity to help companies decide what deserves protection, permission and more investment.

The San Francisco and London company announced a $50M Series B led by Crosspoint Capital Partners on October 14, 2026, bringing reported total funding to $76M. Co-founders Alastair Paterson, CEO, and Bryan Woolgar-O'Neil, CTO, are expanding a platform that applies context to the work employees and AI agents perform.

For enterprise buyers, the interesting commercial possibility is a closer relationship between security and AI adoption. A control that understands the task can potentially inform the next purchasing or training decision, alongside its immediate job of stopping sensitive information from going somewhere it should not.

What Harmonic Security's Series B will support

Harmonic says the new capital will expand engineering and go-to-market operations in the United States and United Kingdom. The company reports more than 90 staff and coverage reaching millions of employees across more than 100 organizations. Those figures describe company-reported reach, with revenue, retention and valuation left undisclosed.

Crosspoint led the financing with participation from existing investors. In a first-party statement reposted by Harmonic, Crosspoint Managing Director Zach Sivertson identified Ten Eleven Ventures, N47 and StepStone Group alongside the lead firm. That gives the investment a more specific institutional context than the announcement's broad reference to returning backers.

The round follows a $17.5M Series A announced in October 2024, led by Next47, now branded N47. Harmonic's earlier release also identified a $7M seed led by Ten Eleven Ventures in October 2023. Its public history places the company’s founding in August 2023, followed by the launch of its data-protection product in July 2024.

That chronology matters because the company now sells into a wider set of workplace behaviors than a browser chat session alone. The latest funding supports a business following AI activity into desktop tools and agent workflows, where an employee's original instruction can lead to subsequent actions across other systems.

Why the task matters to an AI spending decision

Harmonic's Usage Intelligence product groups AI interactions into tasks and business-specific use cases. The company describes views that connect the work to applications, departments and approval status. A security team can examine exposure while a manager considers which tools people actually need.

An unapproved application can therefore become evidence of an unmet requirement. A business might respond by approving the tool, finding an acceptable alternative or helping a team use software it already pays for. That is an operating interpretation of the product design; the value comes from the decision the organization makes with the information.

A procurement conversation also needs evidence beyond activity volume. Frequent use can reflect valuable work, awkward workarounds or a task that still requires substantial correction. Grouping interactions into a recognizable business use case gives the manager a starting point for examining the result, the people involved and the cost of continuing it.

Harmonic's commercial opportunity sits inside that handoff. Security has a reason to understand a workflow before restricting it, and the adoption team has a reason to preserve the workflows that help people deliver. A supplier serving both discussions could become useful earlier in an AI rollout, while permissions, training and spending are still being worked out.

Agent controls bring the decision closer to the action

Harmonic's agent-security materials describe visibility into Model Context Protocol clients and servers, interaction logging and policies governing individual tool calls. MCP connects AI systems to tools and data. Examining the call gives a control a chance to assess what an agent is attempting at the point where the instruction reaches another system.

The company also describes feedback that explains a blocked action to the agent. That is a consequential design choice for an enterprise trying to keep a workflow moving: an explanation can give the software information it needs to attempt an allowed path. The quality of that response depends on the classification, the customer's policy and the available alternative.

This product description sets a clear buying question without requiring a claim of universal protection. The organization needs to understand what is observed, which actions are governed and what evidence remains after an interruption. A successful demonstration should make the permitted work and the constrained action equally understandable to the people responsible for the system.

DevCuration's coverage of Reach Security's control-intelligence expansion examines an adjacent operating concern: how current information about defensive controls reaches the workflow that needs it. Harmonic approaches a different part of that environment, focusing on AI activity and the business task behind it.

The organizational trust behind workforce AI visibility

Understanding work requires access to information about work. Harmonic's usage materials describe task summaries, role-based access controls and options for masking or excluding employees from reporting. These are relevant product provisions for a buyer deciding who should see AI activity, how much detail they need and how the resulting information will be used.

Those choices deserve attention during implementation. Security, the team deploying AI and the employees doing the work may need different views of the same activity. A product that informs both risk management and adoption spending brings those interests closer together, increasing the importance of clearly defined access and purpose.

The current leadership record places Paterson and Woolgar-O'Neil at the center of building that capability. Their expanded engineering team has to maintain useful context across changing AI surfaces, while the commercial team explains where the platform fits alongside controls a customer already owns. Funding accelerates the resources available for both jobs.

At the next renewal, a manager could arrive with a clearer account of why employees keep reaching for a particular tool. Harmonic's opportunity continues through the conversation that follows: who gets permission to use it, what information can travel with the work and how the company helps the next team adopt it.