product-updateproduct-update 6 min read

Google Cloud Opens Gemini Enterprise for Legal and Financial Services

Google Cloud CEO Thomas Kurian published two 25 August 2026 posts putting Gemini Enterprise for Legal and Gemini Enterprise for Financial Services into preview, with a Google-managed Financial Research agent that ships more than 50 skills.

PC

PromptCrates Editorial

Staff Writer

0 0
Google Cloud Opens Gemini Enterprise for Legal and Financial Services

Google Cloud CEO Thomas Kurian published two 25 August 2026 posts putting Gemini Enterprise for Legal and Gemini Enterprise for Financial Services into preview. The finance pack's core is a Google-managed Financial Research agent that ships more than 50 skills.

What Gemini Enterprise is packaging in these two previews

These are the first packaged industry solutions on the Gemini Enterprise platform. They are not a new foundation model and not a new seat SKU in the Antigravity subscription story. Both packs are in preview. Healthcare, Life Sciences, and other professional-services packs are described as next.

Kurian's frame is the same in both posts: a general-purpose model is necessary and not sufficient. Each pack is sold as four pieces that only work together — domain skills, Model Context Protocol connectors that inherit existing permissions, agents that complete work rather than return a suggestion, and an open partner ecosystem — with a governed control plane underneath. The control plane is described as a single dashboard for IT and risk that enforces VPC and CMEK policies, keeps data isolated, and holds outputs to grounded citations.

Client data, playbooks, intellectual property, custom agents, and outputs stay private and are not used to train Google foundation models. That sentence appears in both posts. Treat it as a training-data claim, not as a global privilege opinion.

If your enterprise owner already tracks the IBM and OpenAI consulting partnership around GPT-5.6, keep the two vendor motions apart. IBM-OpenAI is a services overlay on OpenAI models. Gemini Enterprise for Legal and Financial Services is Google Cloud packaging agents and connectors on its own platform. One is a consulting alliance. The other is two vertical previews.

The Legal pack ships purpose-built skills for contract review and redlining, playbook creation, regulatory horizon scanning, legal research, DSAR fulfillment, motions-to-seal redaction, and NDA drafting. MCP connectors inherit existing permissions rather than flattening ethical walls. Access is bound by the role-based and document-level controls already sitting in the document-management and ediscovery systems.

Named connectors include Google Workspace, Microsoft 365, iManage, NetDocuments, Docusign, Everlaw, RelativityOne, Thomson Reuters HighQ, CourtListener, Harvey, Solve Intelligence, and Legora. Systems-integrator partners named on the Legal post include Accenture, Deloitte, KPMG, Devoteam, Factor Law, Tribe.ai, Valtech, Zazmic, Zencore, and 66degrees.

Law firms named as early collaborators are Cleary Gottlieb, Freshfields, Weil, and Williams and Connolly. Their quotes stay inside workflow and governance. Jeff Karp at Cleary talked about slotting Gemini Enterprise into daily tools. Alan Mason at Freshfields pointed at a multi-year Google Cloud partnership. Joe Petrosinelli at Williams and Connolly talked about supporting litigators. Ramona Nee at Weil talked about early access and helping shape the platform.

Do not turn those names into a production-reference list. They are collaborators on a preview. Google did not publish seat counts, matter counts, or a price.

If your team already routes coding work through GPT-5.6 Sol, Terra, and Luna in AWS Kiro, do not put a motions-to-seal redaction skill in that picker. Kiro is a coding-agent surface. This pack is legal DMS connectors and playbook-bound review.

What the Financial Services pack claims it can run

The Financial Services pack is aimed at capital markets and corporate banking. Core is a Google-built, Google-managed Financial Research agent with more than 50 foundational skills. It exposes confidence scores, methodologies, data snapshots, and source citations. Analysts can use it in the Gemini Enterprise app or via Agent-to-Agent APIs.

PR Newswire listed 13 connectors: CoinDesk Data and Indices, Daloopa, Dun and Bradstreet, FactSet, Finnhub, Fiscal.ai, Guidepoint, LSEG, Moody's, MSCI, PitchBook, S&P Global, and SEC Edgar. Google Cloud's own post also names Google Workspace and Microsoft 365 for docs and models. Licensed data stays licensed. Permissioned data stays permissioned. Do not flatten those 13 market-data pipes into "Gemini can see the street."

Named workflows include advisor insights, KYC research that maps corporate hierarchies and ultimate beneficial owners, credit-market idea generation, and bond-issuance pitch decks compressed from days to minutes. Bond-portfolio risk analysis is claimed to drop to a sub-5-minute run, with automated duration-hedging suggestions. Sub-5-minute is the only timing number in the pack. Treat it as a vendor claim on one workflow, not a latency SLA for the agent.

Deutsche Bank is a design partner for the Financial Research agent. Marie-Jeanne Deverdun, the bank's chief technology, data and innovation officer, said the bank will start in the Corporate Bank. CME Group is also named. Broader Gemini Enterprise users cited include BNY, Citi Wealth, Lloyds Banking Group, Macquarie Bank, and Signal Iduna. Those last five are cited as Gemini Enterprise users, not as design partners on this agent.

If you already watch Keenable's $26 million web-search index for AI agents, keep a public web index apart from FactSet, LSEG, and SEC Edgar entitlements. One is a crawl other agents can query. The other is licensed market data inside a customer's environment.

Put the preview status, the 50-plus skills, the 13 finance connectors, and the training-data sentence in the wrapper. A note that only says "Google launched industry Gemini" will collapse two packs, two partner lists, and a design-partner bank into one slogan.

Pin these facts: Thomas Kurian posted both packs on 25 August 2026; both are in preview on Gemini Enterprise; Healthcare, Life Sciences, and other professional-services packs are next; Legal skills cover contract review and redlining, playbooks, horizon scanning, legal research, DSARs, motions-to-seal redaction, and NDA drafting; Legal connectors inherit existing permissions; named Legal collaborators are Cleary Gottlieb, Freshfields, Weil, and Williams and Connolly; Financial Services is aimed at capital markets and corporate banking; the Financial Research agent is Google-built and Google-managed with more than 50 skills, confidence scores, methodologies, snapshots, and citations; usable in the Gemini Enterprise app or via A2A APIs; PR Newswire listed 13 market-data connectors including FactSet, LSEG, Moody's, MSCI, PitchBook, S&P Global, and SEC Edgar; bond-portfolio risk analysis is claimed as a sub-5-minute run; Deutsche Bank is a design partner and will start in the Corporate Bank; CME Group is named; BNY, Citi Wealth, Lloyds, Macquarie Bank, and Signal Iduna are cited as broader Gemini Enterprise users; customer data is not used to train Google foundation models.

Do not write a price. Do not write a general-availability date. Do not write a matter-count or a desk-count Google did not publish. The useful changelog line is two vertical previews, a 50-plus-skill research agent, inherited permissions, and a training-data firewall.

Sources

Google CloudThomas KurianGemini EnterpriseLegalFinancial ServicesDeutsche BankCME GroupFactSetLSEGHarvey

Related articles