Since February 2025, Anthropic has launched 13 vertical AI products and has stopped consistently warning business partners before releasing tools that compete with them.
The shift has hit design and enterprise software companies hard. Figma is down about 50% year to date, raising fresh questions about how AI model providers will coexist with the companies building products on top of their models.
The issue is familiar in tech: a platform company moves up the stack and enters its customers’ market. What makes Anthropic different is the speed of the expansion and the leverage it holds over the model layer.
Anthropic moves deeper into its customers’ markets
Anthropic introduced Claude Code in February 2025 and kept expanding from there. By May 2026, the company had added agents for financial services, Claude for Legal, and Claude for Small Business, according to PYMNTS.
Each product puts Anthropic closer to startups and traditional software companies that have built tools around its models. Those companies now face a supplier that can also become a direct competitor.
Partners say the warning signs are fading
Anthropic previously notified some clients before launching products that could compete with them. That practice is now less consistent, according to The Information.
The report said Anthropic does not always tell partners before entering their markets. It also pointed to the company’s Mythos AI model, which reportedly reduces performance when customers use it to build competing AI software or hardware.
For companies that chose Anthropic as their foundation model provider, that combination narrows the runway. They are not only competing with Anthropic’s products; they are also dependent on Anthropic’s infrastructure.
Figma becomes the clearest casualty
The market reaction was most visible on April 17, when Anthropic launched Claude Design, a conversational tool powered by Claude Opus 4.7. The product can generate prototypes, slide decks, and marketing assets from text prompts.
Figma shares fell 7.5% that day, closing at $18.84 from a prior close of $20.32, according to the Times of India. Adobe also dropped more than 1%.
Three days before the launch, Anthropic chief product officer Mike Krieger resigned from Figma’s board, a seat he had held for about a year, The Information reported. Krieger has publicly argued that the largest AI labs will come to dominate software businesses, according to SmarterX.
Figma has now fallen sharply from its 52-week high of $142.92, reached in August 2025, according to AI Business Weekly. The company went public in July 2025, with shares surging 250% on the first day of trading.
Less than a year later, it is caught in what Wall Street has called the “SaaSpocalypse,” a broader software selloff that has wiped close to $1 trillion from the sector in early 2026.
Platform risk returns, but with more control
Platform companies competing with their partners is not new. Amazon Web Services spent years assuring customers it would not move aggressively into application-layer software, then launched services that challenged companies such as Elastic, MongoDB, and Datadog.
Apple’s annual developer conference has long carried the threat of “Sherlocking,” a term coined after Apple copied key features from Watson, a popular third-party search tool, and built similar functionality into macOS. Google has also displaced partners by bundling maps, email, and productivity tools into its own ecosystem.
The AI boom adds a more serious twist. Model providers do not just control distribution or access to an app store. They control the inference layer.
This provides Anthropic with a much higher degree of control than early platform companies enjoyed. A model can be degraded with respect to specific competitive use cases; thus, the platform provider can degrade the underlying technology that forms the partner’s product.
Software companies look for ways around the threat
Based on The Information, firms that may feel the heat from Anthropic’s expansion fall into one of three categories. They are developing more specialized products for different sectors, emphasizing the compliance and observability features that a generic AI system lacks, and allowing users to select their preferred AI models.
This last point might be the most crucial. In Feb of ’26, Figma themselves partnered with Anthropic on “Code to Canvas”-an integration that “transforms Claude Code responses into Figma designs,” AI Business Weekly reports.
It demonstrates the problem software businesses are in. They must integrate with frontier AIs in order to keep their product relevant even if they risk killing their business themselves by integrating with those AIs.
AI uncertainty freezes software dealmaking
The pressure is also showing up in software mergers and acquisitions. Software deal activity fell to $50 billion in the first five months of 2026, down from $88 billion during the same period last year, according to PYMNTS. That marks the lowest January-to-May deal volume since the pandemic era.
PYMNTS cited industry executives who said buyers are struggling to value software companies while AI reshapes the sector. The concern is simple: many software firms may not be as defensible if model providers can build similar products faster and cheaper.
For Anthropic, the move into vertical software makes business sense. Selling full products can offer better margins than selling API access, especially as the company works to grow revenue before a possible IPO.
For the software companies that built on Anthropic’s models, the equation has changed. Their platform provider is no longer just a supplier. It is becoming a competitor.
