Products & Tools

Genetec's free MCP connector puts Copilot, Claude and ChatGPT to work on security setup

Genetec's free Security Centre MCP connector, released November 19, 2026, links Copilot, Claude and ChatGPT to physical security workflows, with administrators approving every consequential action.

Genetec launches AI connector for security workflows
Genetec launches AI connector for security workflowsgeraldoswald62 / pixabay
By James Calloway5 min read

Why it matters

  • Genetec launched the Security Centre MCP connector on November 19, 2026, free to all customers globally.
  • The connector links Microsoft Copilot, Anthropic's Claude and OpenAI's ChatGPT to Security Centre.
  • Genetec serves more than 42,500 customers across 159 countries.
  • The connector blocks direct database access, so AI agents cannot independently create, modify or delete system data.
  • Users must approve consequential actions before an assistant carries them out.

Genetec has launched a free connector that lets AI assistants such as Microsoft Copilot, Anthropic's Claude and OpenAI's ChatGPT operate inside its Security Centre physical security platform. The Security Centre MCP connector, released November 19, 2026, is available globally to Genetec customers at no extra cost.

The launch targets a specific pain point: repetitive deployment and configuration work in large-scale security environments. Genetec says the connector links AI assistants to Security Centre through MCP (Model Context Protocol, a standard way for AI tools to interact with software), while keeping administrators in charge of what those assistants can see and do.

The market context matters here. Physical security software vendors are under pressure to show AI features with direct operational uses rather than broad automation claims. Genetec has chosen a deliberately narrow entry point: configuration, maintenance and routine administrative work inside an existing security management system, not autonomous control of doors or cameras.

What can the AI assistant actually do?

The first release includes predefined skills aimed at administrators and systems integrators during project rollouts. The tasks include:

  • Applying configuration templates across large numbers of devices
  • Validating site configurations before commissioning
  • Checking connectivity and device health
  • Identifying equipment that needs attention
  • Standardising deployments across multiple sites

Users trigger these actions through prompts to a connected assistant. Genetec gave one concrete example: uploading a new floor plan and instructing the AI assistant to create a new area, along with related sub-areas and doors, without altering existing configurations. Work that might take hours per element could drop to minutes through a single natural-language request, according to the company.

That claim carries the usual caveat. Genetec's hours-to-minutes framing covers a specific configuration task, not the full scope of a security rollout, and the company has not published benchmark figures for how the connector performs across varied deployments.

The takeaway: the connector automates grunt work in system setup, and Genetec's evidence so far is illustrative rather than measured.

How much control do administrators keep?

Genetec has built the connector around a permissions model rather than open access. Administrators decide which MCP permissions and privileges an assistant receives, and whether it can only view information or also carry out actions.

The design has two structural limits. First, the connector uses predefined, outcome-based skills instead of unrestricted access to software development kits or underlying systems. Second, it does not provide direct access to underlying database operations that would allow an AI agent to independently create, modify or delete system data.

In practice, that means an AI assistant can collect information, flag issues or prepare a configuration change, while users remain responsible for approving consequential actions before they happen. The structure is intended to limit risk and make results more predictable, according to Genetec.

The constraint is not incidental. Customers at large enterprises, public institutions and critical sites often demand tight control over what software can do automatically, particularly where systems govern doors, cameras, intrusion tools and other physical security infrastructure. Security vendors adding AI to sensitive products face exactly this question, and Genetec's answer is to keep a human in the approval loop.

The takeaway: the connector cannot change system data on its own, and every consequential action still requires a user's sign-off.

Why does this matter for the security market?

Security Centre unifies video, access control, intrusion, intercom and IoT sensor data on one platform. Genetec says that unified structure provides the basis for broader AI-assisted workflows as customers expand their use of the connector over time. The current release covers deployment tasks only; the company has not specified dates or scope for the next phase.

Genetec, based in Montreal, serves more than 42,500 customers across 159 countries through channel partners and consultants. The company is best known for video management, access control and automatic number plate recognition software. The connector adds an AI integration layer to that business and signals how physical security vendors are responding to customer demand for AI tools that fit within existing systems rather than replace them.

"New technologies create the most value when they strengthen the systems customers already rely on," said Francis Lachance, Senior Director, Product, Genetec. "With the MCP connector, organisations can connect their preferred AI assistant to Security Centre and incorporate it into their security workflows, building on the openness, flexibility, and control of the Genetec platform."

The quote is a positioning statement, not a measurement. What it reveals is Genetec's bet: customers want their chosen assistant wired into tools they already run, with one standard interface replacing separate integrations for each AI product.

The takeaway: Genetec is selling compatibility and control, not AI capability itself.

What is not said?

Three gaps stand out. Genetec has not named which specific skills ship in the first release beyond the deployment categories listed. It has not detailed how the connector audits assistant-initiated actions, beyond the approval requirement for consequential changes. And it has not said whether the free pricing extends to future, broader workflow capabilities — a relevant question for customers planning multi-year adoption.

The connector also depends on third-party AI assistants whose behaviour Genetec does not control. The skills-based design limits blast radius, but customers evaluating the tool will want to know how errors in an assistant's interpretation of a prompt are caught before a user approves the result.

The takeaway: the control model is well-defined at the permissions level, and the operational safeguards beyond user approval remain undescribed.

What comes next?

Genetec frames the connector as a foundation. As customers expand their use, the company expects the unified Security Centre platform to support broader AI-assisted workflows over time. The immediate, verifiable benefit is narrower: faster, standardised rollouts across large device fleets and multiple sites, with no additional licence cost.

For security teams weighing AI adoption, the launch offers a low-risk entry point — read-only or action-capable permissions, human approval gates, and no direct database access for the agent. Whether Genetec turns that foundation into genuinely broader automation will depend on how the skills catalogue grows and whether customers in sensitive sectors actually grant assistants action privileges rather than view-only access.

Source: itbrief.com.au

Share this article:

More from James Calloway

James Calloway

Show full bio

News editor covering people, launches and day-to-day practice in artificial intelligence and machine learning at AI In Context.

749 articles

Related articles

  1. Coalition plugs OpenAI, Anthropic, Gemini and Cursor into Wirespeed
  2. A1 launches Guardrails for mid-market AI governance
  3. Ivanti Opens Neurons Data to AI Clients via MCP, Governed by Existing Permissions
  4. OpenAI Rolls Out GPT-5.4-Cyber to Vetted Defenders
  5. 87% of Enterprise Developers Use Low-Code. That Is the Agent Security Problem

« Previous articleNext article »