Docusign held its annual conference this week in New York and announced a set of tools that together allow AI assistants to do real work on your contracts, not just answer general questions about them.
The most immediately relevant piece for non-technical professionals is the MCP Server, now in open beta globally. MCP, which stands for Model Context Protocol, is a standard connection layer that lets AI tools talk to external software systems. Anthropic created it in late 2024 and it has since become the common way most AI platforms connect to business tools. In plain terms: it is the reason you can soon ask Claude "show me all contracts with this supplier" and get your actual data back, not a generic AI response.
Docusign's MCP Server connects to Claude, Gemini, Microsoft Copilot, and GitHub Copilot. A user authenticates with their Docusign account, and the AI assistant can then check contract status, search the agreement repository, trigger approval workflows, and send signature requests, all through typed or spoken instructions. Permissions stay tied to whatever the user is already authorized to see in Docusign, so the AI cannot access anything the person could not access manually.
The second major launch is Agent Studio, currently in early access. This is a builder tool that lets organizations create AI agents trained on their own contract history, internal policies, and playbooks. An agent built in Agent Studio could, for example, automatically flag renewals, identify risk clauses based on the company's own standards, or route a contract through an approval chain when certain conditions are met. The agents run inside Docusign's platform and every action they take is logged, with the option to require human sign-off before the agent proceeds.
The third piece worth understanding is the Agreement Manager API. This is what makes bulk contract ingestion possible: pulling thousands of contracts from Salesforce, SharePoint, Google Drive, or legacy archives into a single searchable repository. Once documents are in, Docusign's AI reading system called Iris analyzes each one and extracts structured data: expiry dates, obligations, counterparty names, clause history, past negotiation positions. That extracted data is what agents and AI assistants draw from when they answer questions or take action.
For context on scale: Docusign's intelligent agreement platform already accounts for over $350 million in annual recurring revenue and is growing fast as a share of the company's total business. The company serves over 1.8 million customers globally, meaning the infrastructure behind these new tools is not experimental, it is running at enterprise scale.
The practical implication for business operators is straightforward. If your organization manages a large volume of contracts, including supplier agreements, employment contracts, NDAs, or client MSAs, this combination of tools means your existing AI assistant can soon become an active part of contract management rather than a general-purpose search tool. The prerequisite is that your contracts need to be in, or migrated to, Docusign's repository. Companies that do that groundwork now will be positioned to get real value from AI-assisted contract review, renewal tracking, and risk flagging much faster than those who wait.