| published by | Amanda Caswell |
|---|---|
| in blog | The New Stack |
| published date | 2026-11-14 |
| original entry | Anthropic wants you to build on Claude. There’s a catch. |
Anthropic wants developers to build businesses on Claude, but its expansion into enterprise software is raising concerns among founders that the company supplying their models could eventually become their biggest competitor.
Gerald Zingraf, president and COO of GlacierGrid, raised the concern on LinkedIn following Anthropic’s Tuesday expansion of Claude’s enterprise lineup, comparing its strategy to Amazon’s private-label business.
“Makes me nervous for wrappers and pure SaaS companies,” Zingraf writes, drawing on his experience selling products he designed through Amazon before watching the retailer introduce competing products.
“Makes me nervous for wrappers and pure SaaS companies.”
His concern is that Anthropic can see which third-party integrations customers use most frequently and potentially draw on that information when deciding which capabilities to build itself.
“Anthropic has the same view,” he argues, although he offers no evidence that the company uses integration data to identify competitors or inform its product roadmap. For some developers, the risk is no longer hypothetical.
The New Stack reached out to Anthropic to ask whether the company uses data from third-party integrations, including connector usage and tool invocation frequency, to inform its product roadmap. We also asked what visibility Anthropic has into how customers use third-party tools, whether safeguards prevent that information from being used to develop competing native features, and how the company plans to balance its relationships with SaaS partners against its expansion into their markets. Anthropic has not responded as of publication.
Scott Stevenson, CEO of legal AI startup Spellbook, described what happened when Anthropic moved into his company’s market in a November 2 interview with Forbes.
Spellbook uses Claude models for everything from reviewing contracts and suggesting revisions to automating legal workflows. Stevenson recalls Anthropic employees assuring him the company had no plans to compete directly in legal software, before Anthropic launched Claude for Legal, which includes contract-review capabilities that overlap with Spellbook’s business.
“Anthropic has been a great partner, but it made us realize that they are a massive company and two parts of the org can have conflicting mandates,”
“Anthropic has been a great partner, but it made us realize that they are a massive company and two parts of the org can have conflicting mandates,” Stevenson said in the interview. “Ultimately this is inevitable.”
Spellbook hasn’t abandoned Claude. Anthropic still features the startup as a customer, reporting roughly 530,000 monthly contract reviews performed by its Claude-powered agents, while Stevenson also appears in Anthropic’s October 28 announcement of legal artifacts.
The relationship illustrates a difficult position for developers whose products depend on frontier models. Spellbook still benefits from Anthropic’s technology, even as Anthropic develops applications aimed at some of the same customers. Nothing in Stevenson’s account establishes that Anthropic used Spellbook’s proprietary information to develop its legal offering.
Tech founder and Dalton Mills CEO Todd Saunders made a similar argument in a March 10 post on X, following Anthropic’s introduction of Claude Code Review.
“Anthropic is the new Amazon.”
“Anthropic is the new Amazon,” Saunders writes, warning developers building on Claude’s SDK that the company could observe application usage and eventually incorporate competing functionality into Claude.
The company’s announcement earlier this week introduced Claude Sheets, Claude Whiteboards, and Claude Data, while previewing Claude Meetings. Expanded interactive connectors also let partners embed applications directly inside Claude, including software from Adobe and Atlassian.
Anthropic has access to information about how customers interact with third-party apps and services, something the company acknowledges in its privacy policy, effective September 10, 2026. What isn’t clear is whether Anthropic looks at which integrations are most popular when deciding what to build next.
The rules are different for business customers. Under Anthropic’s commercial terms, customers retain rights to their inputs and own their outputs, while Anthropic agrees not to use their content to train its models. Its commercial privacy documentation also says feedback data excludes content pulled through connectors, including local and remote MCP servers, unless users paste it directly into a conversation.
Neither document settles how Anthropic uses aggregated integration data.
The company is still encouraging developers to build on Claude. It’s just that access to its growing enterprise customer base comes with a tradeoff, since Anthropic controls the interface and can introduce competing features without consulting the developers who depend on it.
For now, Anthropic is asking developers to trust it as both a technology provider and a business partner, even as it moves into markets those same developers are trying to build businesses in. Whether that relationship remains sustainable may depend less on what Anthropic promises today than on what it decides to ship next.
AI Engineer Step by step guide to becoming an AI Engineer in 2026 Introduction Working With LLMs Context Engineering AI Models Embeddings & Vector Databases Embedding Models RAGs AI Agents AI Safety and Ethics Evaluation & ObservabiityYOUTUBE.COM/THENEWSTACK
Tech moves fast, don't miss an episode. Subscribe to our YouTube channel to stream all our podcasts, interviews, demos, and more.