Anthropic vs. the lawyers: why service-provider stocks are in free fall

The stock market is rarely merciful to those who fall behind progress, but on Tuesday we watched a textbook public flogging. As soon as the startup Anthropic announced the launch of its new tool for the legal sector, the share prices of companies that provide data-related services sank in unison. This is the moment when theoretical talk about AI replacing people turned into concrete losses in investors' portfolios. We are watching the beginning of the end for a business model that fed entire corporations for decades: selling person-hours for boring office work.

Anthropic did not simply update its Claude model. It built a specialized solution that can analyze legal documents, spot risks and cross-check facts with the effectiveness of an experienced lawyer. For the business process outsourcing industry, this sounds like a death sentence. Why pay a huge staff in India or Eastern Europe for a first-pass contract audit if an algorithm does it in seconds, at practically no cost? Investors priced in this risk immediately, understanding that the margins of classic service providers will soon head toward zero.

The roots of this situation go back to last year, when shares of the education service Chegg collapsed after it admitted that students preferred ChatGPT to its services. Now the story is repeating itself, but on a more serious, corporate scale. If AI was once seen as an amusing assistant that can write a letter or draw up a workout plan, it is now moving into the territory of complex expertise. Legal analysis is only the first swallow. Financial auditing, consulting and deep data analytics are next, fields where the "human factor" was previously considered irreplaceable.

Why is this happening right now? Anthropic and its competitors at OpenAI have realized that general-purpose chatbots are just the tip of the iceberg. The real money lies in building narrowly focused agents that integrate into companies' workflows. When AI stops merely "talking" and starts "doing" specific work, the value of the intermediaries who used to provide that service with human labor drops to zero. We are seeing a shift from the economy of cloud computing to the economy of intelligent action.

Many executives at data-processing companies are trying to put a brave face on things, saying they are adopting AI inside their own organizations. The market does not believe them. The problem is that if you adopt AI to do the work ten times faster, you can no longer bill for the time spent. This is a fundamental conflict of interest that is breaking the old services economy. While service providers try to restructure, startups like Anthropic simply take their bread by offering direct access to the technology.

In the coming months we will probably see several more such crashes in other sectors. Any company whose main asset is "people at computers following instructions" is at risk. Investors will now look not at revenue but at how well a business model is protected from the direct impact of language models. The irony is that the very data these companies collected and processed for years has ended up as fuel for training their digital killers.

The bottom line: Anthropic has shown that specialized AI agents are ready to kill entire service markets. Who is next in line to clear out their desks: accountants or junior analysts at investment banks?