Autologyx today (15 June) announced a new set of MCP-enabled capabilities designed to help law firms and other regulated organisations move from AI experimentation to governed AI execution.
The announcement introduces support for approved AI agents to participate directly in workflows running on the Autologyx platform. Rather than operating as standalone assistants outside the business process, agents can work within Autologyx to support activities such as updating records, creating tasks, generating documents, progressing matters and assisting users with operational work – within a controlled and auditable environment, which is absolutely key. Agents operate within the platform’s permission model and can be controlled at a granular level based on user, group, process, matter, client, data set or workflow stage. Organisations can determine not only what a user can access, but also what an agent and model is acting on their behalf is permitted to see or do.
The launch reflects a broader shift taking place across the legal industry as organisations move beyond evaluating AI tools and begin addressing the operational, governance and accountability challenges associated with deploying AI at scale.
“Connecting AI agents to systems is the easy part,” said Ben Stoneham, founder and chief technology officer at Autologyx. “The real challenge is governing what happens once they are connected. Organisations need to know what an agent can access, what actions it can perform, who authorised those actions, and how those decisions can be audited and explained. That’s where we believe the market is heading.” This accords with the conclusion of Legal IT Insider’s brand new GenAI report out tomorrow (see sign up details below), which Autologyx features in extensively.
The new capabilities build on Autologyx’s existing governance and orchestration architecture, which already supports complex operational workflows across law firms, financial services organisations and other highly regulated environments.
“Many firms have invested heavily in AI pilots and point solutions over the last two years,” said Stoneham. “What we’re increasingly hearing from customers is that the next challenge isn’t access to AI. It’s operationalising AI safely. They need a governance layer that sits above individual tools and provides visibility, accountability and control across the work being performed.”
Alongside the new MCP capabilities, Autologyx is introducing enhanced audit and monitoring functionality designed to provide a more complete view of human and AI activity across workflows. The enhanced audit capabilities provide visibility into interactions, context, model outputs, tool usage, workflow activity and resulting actions, enabling organisations to understand not only what happened, but how and why actions were taken. The capability is designed to support firms that require stronger oversight of AI-assisted work in client-facing and regulated processes.
Every action performed within Autologyx remains linked to the relevant matter, process, document, record or datapoint, creating a unified timeline of activity that can be reviewed by operational, risk, compliance and legal teams. The company says this governance-first approach will become increasingly important as organisations move towards more autonomous forms of agentic work.
“Today, most AI deployments still involve significant human oversight,” said Stoneham. “As agents become more capable and more integrated into business processes, organisations will need stronger mechanisms for attribution, auditability and control. The ability to understand which agent performed an action, under what authority, against which data, and with what outcome, becomes fundamental.”
The new functionality will be showcased at LegalTechTalk 2026.
To sign up for our brand new GenAI report out tomorrow, click HERE
