Google Cloud on Tuesday launched two industry-specific versions of Gemini Enterprise for legal and financial services.
Gemini Enterprise for Legal uses purpose-built skills for tasks including contract review, legal research, regulatory monitoring and data access requests. It also uses secure Model Context Protocol (MCP) connectors to connect AI agents with legal systems while retaining existing permissions and access controls.
The platform includes agents for legal research, regulatory screening and contract drafting.
For financial services, Google has introduced a Financial Research agent with more than 50 foundational skills. It provides confidence scores, methodologies, auditable data snapshots and source citations, and can connect to enterprise data through MCP and other agent workflows through Agent-to-Agent (A2A) APIs.
The financial-services offering targets KYC research, portfolio-risk analysis, credit-market research and bond issuance.
The offerings bring AI agents into workflows across law, capital markets and corporate banking.
“Industry context is the foundation. Enterprise context is the advantage,” Google Cloud said.
According to Google, a general-purpose model can analyze a financial filing or summarize a contract, but it does not automatically understand an institution’s business logic, risk appetite or permission boundaries. Gemini Enterprise for Legal and Financial Services is built to work within these industry-specific requirements.
For legal teams, the system is designed to apply an organization’s standards and institutional knowledge while identifying where expert judgment is required.
Tribe AI has joined Google Cloud’s partner ecosystem for Gemini Enterprise for Legal and Financial Services as an implementation partner, helping enterprises deploy the technology within their existing operations.
“We connect Google’s foundation to each company’s data, systems, workflows, and controls, then take it all the way into production in a way that fits how the business actually operates,” said Jaclyn Rice Nelson, co-founder and CEO of Tribe AI.
Google claims that customer data, business rules, intellectual property, custom agents and model outputs remain private and are not used to train or fine-tune its foundation models. Both offerings are available in preview.
Also Read: What are the “Most Urgent AI Risks” according to 272 Experts?






