Anthropic Is Blindsiding Its Own Business Partners, The Information Reports

Anthropic Is Blindsiding Its Own Business Partners, The Information Reports

Saturday, June 14, 2026 | 4 min read | 820 words

Key Takeaway: Anthropic has repeatedly launched products that compete with its own partners — while simultaneously announcing an expanded partner network and moving toward a potential IPO — raising questions about whether it can be trusted as a long-term platform.

---

A new report from The Information paints an uncomfortable portrait of Anthropic's partner relations: the company has a pattern of inviting firms into high-profile launch events while withholding the most consequential detail — that the product being announced competes directly with them.

The headline case is Claude Design, Anthropic's AI tool for generating visual mockups, UI prototypes, and presentation decks, which the company unveiled in April 2026. Weeks before the announcement, Anthropic approached several companies — including Figma and Canva — asking them to participate as "partners" of the launch. Both are design-software companies whose core products overlap with what Claude Design does. Figma's share price fell roughly 5% on launch day, three days after Anthropic's own chief product officer, Mike Krieger, had quietly resigned from Figma's board of directors. Canva fared differently: it is the only firm that appears to have co-developed Claude Design with Anthropic, and Claude Design shipped with a dedicated "export to Canva" button — a preferential integration not afforded to anyone else.

The result was an asymmetric partnership: one design-software company was woven into the product, while another learned the full scope of what it had agreed to promote only after the announcement was made.

The Pricing Shock

A separate episode, also surfaced in secondary reporting, concerns Claude Enterprise customers. In April 2026, Anthropic shifted those customers from a flat per-seat fee — historically as high as $200 per user per month — to a usage-based billing model. Under the new structure, companies pay a $20-per-user monthly base fee plus additional charges tied to actual compute consumption. The transition was made with little advance warning.

Fredrik Filipsson, co-founder of software licensing negotiation firm Redress Compliance, estimated the change could double or triple costs for heavy users. "Weekly active users of Claude Code doubled between January and February 2026," he noted, pointing to the kind of compounding consumption that makes a token-based model dramatically more expensive than a flat rate. Anthropic said the pricing change did not apply to businesses paying for fewer than 150 users, limiting the blast radius — but for large enterprise deployments, the math changed overnight.

A Familiar Tension in Platform Strategy

Neither episode is unique to Anthropic. Large technology platform companies have long faced accusations of "platform risk" — adopting third-party use cases as first-party products, sometimes after learning which ones gained traction from partners. Amazon has faced it with marketplace sellers. Apple has faced it with App Store developers. Google has faced it with search-adjacent products.

What is different about Anthropic's situation is timing. The company is simultaneously:

- Expanding its Claude Partner Network, into which it has committed $100 million and enrolled major consulting firms including Accenture, Deloitte, Cognizant, and Infosys. - Pursuing a potential IPO as soon as Q4 2026, with a valuation that has reportedly reached $965 billion following a confidential filing. - Trying to establish Claude as the default enterprise AI platform at a moment when Microsoft, Google, and OpenAI are competing for the same contracts.

Enterprise software buyers are not consumers. They sign multi-year contracts, run procurement processes, and — critically — have long institutional memories. A company preparing to go public on the strength of its enterprise revenue base cannot easily afford a reputation for treating partners as convenient billboards for products that undercut them.

Anthropic's Case

To be fair to Anthropic, the company is not operating irrationally. The AI market is moving fast enough that delaying product launches to manage partner sensitivities could mean ceding ground to competitors who have no such constraints. Claude Design entering the design-tool space is a legitimate business decision; how it was communicated is the question.

And the usage-based billing change, while jarring, is arguably more honest than a flat-fee model that obscures the real cost of heavy AI usage. Anthropic's argument — that a $20 base plus consumption pricing better reflects actual cost — is a reasonable one for customers who use Claude lightly. The problem is in the execution: heavy users, who were presumably among Anthropic's most engaged enterprise customers, bore the full shock of the transition.

What to Watch

Whether Anthropic addresses this friction before its IPO will be a signal worth tracking. Public investors will scrutinize customer concentration, contract durability, and churn risk. Partners who feel blindsided do not renew. The Claude Partner Network's $100 million commitment is a credible show of intent — but it lands differently if the partners writing case studies discover they are also the competition.

---

Sources: - [The Information](https://www.theinformation.com/articles/anthropic-blindsides-business-partners) - [PYMNTS](https://www.pymnts.com/artificial-intelligence-2/2026/anthropic-switches-to-usage-based-billing-for-enterprise-customers/) - [VentureBeat](https://venturebeat.com/technology/anthropic-just-launched-claude-design-an-ai-tool-that-turns-prompts-into-prototypes-and-challenges-figma) - [Fast Company](https://www.fastcompany.com/91538439/design-enters-its-frenemies-era) - [Groundy](https://groundy.com/articles/anthropic-ends-flat-fee-enterprise-claude-above-150-seats-and-forces-per-token-billing-on-ai-procurement/)