Anthropic on Friday released a plugin for its Claude Cowork AI agent, enabling automation across legal, sales, marketing, and data analysis. Photo illustration: Jonathan Raa/NurPhotoJonathan Rahr/Reuters
This pattern should be familiar by now. A demonstrable advance in artificial intelligence is expected to transform a traditional industry, causing a panic in the market and a fall in the stock prices of traditional companies.
The data provider and legal software company’s turn came Tuesday after Anthropic released productivity tools for lawyers. Many companies that provide databases, analytics and other tools to the legal industry have suffered in markets around the world as investors have grown increasingly skeptical of software providers of all kinds in the face of AI. The Wall Street Journal reported that two State Street SPDR exchange-traded funds, which track software, services and financial data stocks, lost $300.6 billion in market value following the news.
The stock selloff is similar to what happened last year when Chinese company DeepSeek released a cheap and efficient AI model that drove down the stock prices of companies involved in building resource-intensive data centers. That reaction turned out to be overblown.
Market factors: Worried about asset bubbles? There’s an index for that
There is so much uncertainty about AI winners and losers that we don’t know if that will happen again. Hardware companies such as Celestica Inc., which makes equipment for AI data centers, have performed well over the past year due to demand for their products. Software developers are another story. Toronto-based Constellation Software Inc., which buys companies that build applications for niche industries, has seen its stock fall more than 50% in the past 12 months on concerns that its AI coding tools will make it easier for competitors to win business and for its customers to create custom tools.
Tuesday’s market reaction also highlights tensions between frontier AI developers like Anthropic, OpenAI, and Google, which build general-purpose large-scale language models (LLMs) that can complete a variety of tasks, and companies that build products on top of these very same models.
For example, Anthropic recently released a plugin for its generative AI platform. It says it can handle legal tasks such as document reviews, risk reporting, and compliance tracking. This represents a potential threat not only to established legal tech companies but also to the many AI startups building tools for lawyers.
Thomson Reuters, which has invested heavily in AI, saw its stock price drop 16% on the Toronto Stock Exchange on Tuesday. (Woodbridge, Thomson Reuters’ controlling shareholder, also owns the Globe and Mail.) CS Disco, a U.S. AI-powered legal services company, fell 12%. Meanwhile, LexisNexis owner RELX and Dutch legal analysis service provider Wolters Kluwer fell 15% and 13%, respectively.
The question is: What’s stopping Anthropic and its ilk from moving deeper into the legal tech space, and other sectors for that matter, building custom software based on their model and potentially putting their own customers out of business?
“These applications are just wrappers around what already exists in LLM and don’t understand what a permanent moat is,” said John Ruffolo, founder of Maverix Private Equity. “You’re building your business on top of another business that may be competing with you.” That’s one reason Mavericks has avoided backing AI startups, despite their soaring valuations.
How AI is hurting and helping the next generation of workers
But when it comes to law, there’s a reason AI developers may not put in much effort for the time being, says Scott Stevenson, co-founder of AI legal software company Spellbook. First, the market may not be large enough to be worthwhile, but the expertise some companies offer is not easily replicated. Spellbook, for example, boasts about 4,000 customers who use AI tools for contract reviews.
“Model providers will start building some lightweight tools for different industries, but what Anthropic is doing is very thin,” Stevenson said. “Spellbook is like a toaster. We do one thing and we do it well.” With each new LLM release and as law firms’ interest in specialized tools increases, Spellbook’s business only grows, he said. But as general-purpose AI systems advance, every other company will have to step up its game.
Other sectors may be more vulnerable to disruption. Companies like Cursor, Replit, and Lovable have garnered huge valuations thanks to the popularity of their AI coding tools that allow them to build apps and websites based on instructions in easy-to-understand language. But while these products leverage LLM from cutting-edge developers, Anthropic’s Claude Code has become the de facto platform for many software developers. OpenAI has its own coding platform called Codex.
For these large developers, building coding tools is easier than tackling complex integrations with other software and processes that exist in other industries, and given the size of the software market, it can also be more profitable. Additionally, Stevenson said AI companies are heavy users of these tools, so they have an incentive to keep improving them. “I think they’ll definitely compete with Cursor and other players,” he said.
For investors, it is not entirely clear which companies will be most affected by AI, let alone which sectors. For example, the S&P North American Technology Software Index is down about 19% since the beginning of the year. Shares of video game companies such as Take-Two Interactive Software Inc. and Roblox Corp. plummeted in late January after Google released Project Genie, which allows companies to build digital worlds based on text prompts.
“The competition in AI is probably the most ruthless competition in the history of technology,” Stevenson said. “The only moat now is speed.”
