Why the Market Overreacted to Claude's Legal Skills Announcement

This week's announcement that Anthropic had introduced "legal skills" inside its Cowork environment for Claude triggered an immediate reaction in the public stock markets. Shares of established legal and regulatory information providers, including Thomson Reuters, Reed Elsevier, and Wolters Kluwer, tanked as investors appeared to price in a new competitive threat.

These stocks have already seen declines, suggesting a reset already, and tech stocks were down across the board, too. I feel the market reaction does not reflect the reality of the situation.

The market reacted to a horizontal platform signaling more vertical ambition, which is hard even outside of the legal vertical.

Here are ten reasons why I believe the market overreacted.

1. Cowork Is Not Aimed at Law Firms - And That Matters

Anthropic's Cowork is targeted principally at corporations. Legal is one set of functional skills. There are also financial services skills. And skills exist to support sales, marketing, and HR. While their legal skills appear strong, they are not comprehensive. I wonder whether the inclusion of legal skills is strategic or opportunistic to support enterprise adoption. Anthropic can be successful in enterprises even if the use of legal skills were to be limited.

Cowork's legal skills skew toward transactional and drafting. By contrast, the bulk of Thomson Reuters and LexisNexis usage and revenue comes from law firms and is driven by litigation, legal research, and regulatory compliance. The combination of market segment and use case is a first line of defense.

2. Cowork Focuses on Transactional Workflow

Cowork should pose a greater threat to contract-drafting solutions. Even so, legal tech drafting tools have operated in the backdrop of Microsoft's Word, SharePoint, and CoPilot.

The publicly traded incumbents have some revenue streams from transaction drafting tools, but that isn't the core of their offerings. Cowork could orchestrate workflows, but the skills look like the functions of an assistant for transactional work right now.

3. The Law Department Is Not Just Another Enterprise Function

Legal work is different. AI marketing outputs require a few legal guardrails to ensure compliance, which can be handled through human legal review. Marketing output can be subjective. But Legal work has many compliance and accuracy requirements. It must withstand the challenges posed by counterparties, adversaries, courts, and regulators.

Legal requirements, such as privilege, also create wrinkles in the fabric of horizontal solutions. "Legal wrinkles" are why Microsoft, despite its strong position, has never fully committed to legal. If Microsoft hasn't, will Anthropic or OpenAI be successful?

4. Litigation Is Where the Money Is

Cowork is intended to orchestrate workflow, but the legal skills center on contract drafting and review. There are many other matter types. And when it comes to where the money is, Thomson Reuters and LexisNexis drive most of their value in legal research tied to disputes and litigation. Winning and losing matter. This is why services like Westlaw and LexisNexis will remain essential as they provide legal authority, including comprehensive primary sources with citations to laws, regulations, and caselaw.

5. Legal Is an Ecosystem, Not a Feature

Legal platforms operate within an ecosystem. As long as judges expect West-style citations, including page-specific references, the upper hand goes to the established players. If courts continue to use Westlaw or LexisNexis, then law firms will too.

Cowork currently operates outside that ecosystem.

6. Partial Solutions Don't Replace Systems

Contract lifecycle management (CLM) platforms maintain a moat by acting as the central repository and system of record for agreements. Is Anthropic willing to become the system of record? If not, then Cowork can orchestrate a process, but it can't replace a CLM. Skills can support pre-signature negotiations. If the final document is still stored in a central repository for post-signature reporting and analytics, is that system really at risk?

Without providing a full solution to a vertical problem, Anthropic's orchestration will augment existing solutions rather than displacing other vendors like a CLM.

7. Integration And Redesigned Workflows

Cowork's orchestration and agents can be used to redesign workflows and connect to an organization's proprietary datasets. Real lock-in occurs when engineers map processes and implement integrations. This "forward engineering" is one where specialized players like Harvey and Legora shine. They perform this work and then are around in some capacity after deployment.

Anthropic does have forward-deployed engineers, but will they be able to dedicate legal-specific engineers? Or will they more likely be generalists?

If anything, Anthropic is more likely to compete directly with Harvey and Legora for projects as true legal verticals. Again, Thomson Reuters, LexisNexis, or Wolters Kluwer shouldn't be particularly threatened.

8. The Last Mile Is Also Human

There is a technical aspect to the last mile, but a human one too. Enterprise legal buyers require onboarding, training, and ongoing support. Remember the outrage when Thomson Reuters tried to pull the plug on its 24/7 legal research helpdesk?

The publicly traded incumbents have spent decades building support infrastructure and relationships. Harvey and Legora are doing that too. Will Anthropic add to its cost structure to support the high-touch last mile? That remains to be seen.

9. Commercial Reality Favors Incumbents

Platforms like Westlaw and LexisNexis have large user bases. Lawyers don't like change, and switching costs can pile up when moving away from familiar products.

There are also contractual considerations. The incumbents prefer multiyear contracts with larger organizations. When renewals finally come up, they are heavily negotiated and cover a wide range of products. How hard would it be for a law firm to let go of Law360? Unless an organization looks to terminate an entire vendor relationship, true cost savings may be less than imagined.

As Anthropic's solution matures, it will take years to whittle away at the well-fortified positions of the incumbents.

10. Vertical Specialization Can Be a Distraction

Vertical specialization can run counter to the economics of large-scale platforms. This is why SAP provides limited tax functionality in its financial accounting platform. Tax law changes constantly, and the processes that support ever-changing tax functionality are markedly different from those for static accounting functions.

LLMs could offer an exception to the economics of vertical specialization. But the nuances of legal vertical will be a distraction for those trying to operate at scale. The exception is the consumer market for legal, which is large and less sophisticated.

Conclusions

Anthropic, OpenAI, or Microsoft could win in legal. The question is how and if it is worth the hassle.

Entering legal to win requires sustained investment, perhaps even an acquisition of an incumbent. They will risk being distracted from the simplicity of the horizontal business model. This is why partnering to deliver the last mile can be appealing. It is also why specialized legal tech vendors have long existed and why new ones have emerged.

With continued commitment and expansion, Anthropic's moves could impact the publicly traded incumbents. But a sober look at the obstacles suggests the market reaction is overstated for now.