What happened
Bloomberg Law reports that corporate legal departments and law firms are competing to capture legal knowledge in AI systems. Companies are using software to extract contract clauses, legal advice and other institutional expertise so attorneys can query it later, while firms are automating routine work and emphasizing specialized, high-stakes matters.
Bloomberg Law reports that law firms and corporate legal departments are in a head-to-head effort to capture profitable legal knowledge in systems that use AI to make expertise more accessible. The material at stake includes preferred contract clauses, legal strategies, research and explanations of prior advice. The report says corporate teams such as Salesforce and Palo Alto Networks are using advanced technology to move more work in-house, while firms including Kirkland and Willkie are trying to use specialized knowledge to retain clients. Bloomberg Law quotes legal technology adviser Monica Zent describing knowledge and data as the critical competitive asset. The source does not independently confirm the reported systems’ effectiveness.
According to Bloomberg Law, AI is accelerating a longer-running trend in which corporate legal departments take on more routine work themselves, including legal research, contract review and due diligence. The report says companies such as Ford and Toshiba have pursued that approach to reduce costs. It also describes software companies Summize and Eudia as pitching tools that extract important points from legal work and make them available for future matters. Summize chief executive Tom Dunlop calls the change a “memory shift,” in which institutional knowledge can become repeatable and searchable. Bloomberg Law reports that many customers still lack a sufficiently good database of institutional knowledge to support these systems.
The report gives several examples of organizations attempting to preserve expert knowledge. Bloomberg Law reports that Coinbase is capturing the knowledge of senior experts in generative-AI platforms and that former legal chief Paul Grewal is helping with a transition through Oct. 31 while an AI agent avatar nicknamed “Better Call Paul” is expected to remain available afterward. It also reports that AT&T’s LegalEdge program automates contract review by pairing the system with a contracting expert who breaks down individual clauses. Law firms are described as automating lower-level services as well: Wilson Sonsini has platforms for startup incorporations and nondisclosure agreements, and Cooley has launched chatbots. These details are attributed to Bloomberg Law and are not independently confirmed here.
Read the primary source: news.bloomberglaw.com ↗
Why it matters
The report describes a potentially important shift in the client-law-firm relationship: AI may make knowledge that once remained inside outside counsel more reusable by corporate legal teams. That could move more routine work in-house, but the source also reports that firms retain broader experience across many clients and matters.
The central significance is control over institutional knowledge. Bloomberg Law reports that in-house teams historically depended on law firms to store information, including some companies’ patent records, and did not always receive explanations that could be reused on later problems. If AI systems can reliably organize prior advice and contract patterns, a company could distribute a legal answer across its organization instead of repeatedly asking outside counsel the same question. That would make legal knowledge more portable and could strengthen the bargaining position of corporate clients.
The change could affect which legal tasks remain with outside firms. Bloomberg Law reports that firms such as Cooley and Goodwin are treating “high value” work as a strategic citadel while using technology around the edges. Their proposed advantage is exposure to many comparable matters, whereas an in-house lawyer may see only the disputes or transactions of one company. The source quotes Northwestern law professor Daniel Linna making that distinction. The report therefore does not support a simple conclusion that AI will eliminate outside counsel; it describes a contest over which side can preserve and deploy the more useful knowledge.
There are practical governance implications. A system trained or configured with contracts, litigation strategies, advice and expert judgment could influence future legal decisions across an organization. The source does not establish how the organizations cited handle confidentiality, privilege, access controls, errors, retention or accountability when AI-generated answers are reused. Nor does it provide independent testing showing that these tools give accurate legal guidance. Those unknowns matter because making institutional memory more available can increase efficiency while also making an incorrect or improperly exposed answer easier to repeat.
What to watch next
The key questions are whether companies can assemble reliable internal knowledge bases, how firms and clients will share data, and who will control collaborative systems. Bloomberg Law reports examples of AI agents, contract-review tools and proposed shared cloud spaces, but the source does not independently verify their performance or long-term adoption.
Bloomberg Law reports that the next stage may involve shared cloud systems, vaults or workspaces in which law firms and clients store and query documents together. Meta legal operations chief Mike Haven is quoted saying that ownership of such a shared space may matter less than the ability to collaborate in one place. The unresolved issue is control: the source does not say who would set permissions, maintain the underlying data, determine which advice is authoritative or bear responsibility for an AI-assisted mistake.
Companies’ ability to build usable knowledge bases is another immediate test. Bloomberg Law reports that many organizations still lack the clean, structured institutional knowledge needed to support AI. AT&T’s example suggests that deployment requires experts to identify and explain the reasoning behind individual contract clauses, rather than simply uploading documents. Readers should watch for evidence about implementation scale, accuracy, review requirements, measurable time or cost changes and whether systems work beyond narrow contract-review tasks.
The report also points to a broader organizational question: whether firms and corporate legal departments can develop enough communication to share knowledge without undermining their separate interests. Bloomberg Law reports that adviser Meredith Brown sees firms preserving their core business while adding AI-assisted capabilities, and that Linna identifies insufficient conversation between firms and corporate departments as a problem. The source offers no confirmed timetable for a new industry model, no independently verified adoption figures and no evidence that the reported examples represent the entire legal sector. Those limitations should temper claims about how quickly AI will reshape legal services.
