Anthropic Strikes CoreWeave, Broadcom Deals; Burry Eyes Palantir

A flurry of partnership news and a notable hedge-fund call put Anthropic at the center of AI market conversation this week, with new compute agreements and a high-profile competitive warning against Palantir all landing within days of each other.

Key points

• Anthropic and CoreWeave signed a new infrastructure agreement, with CoreWeave’s AI compute capacity set to power Anthropic’s models; CoreWeave stock surged 13% on the news.
• Broadcom announced chip deals with both Google and Anthropic, sending Broadcom’s stock higher on Tuesday.
• Famed short-seller Michael Burry publicly positioned Anthropic as a serious threat to Palantir, citing enterprise ease-of-integration as a key differentiator.
• The deals add to Anthropic’s growing web of compute and silicon partnerships, which investors tracking a potential IPO will likely view as evidence of scaling intent.

What do the CoreWeave and Broadcom deals actually signal?

The CoreWeave agreement is the more immediately market-moving of the two. A 13% single-day pop in CoreWeave stock suggests investors read the deal as a meaningful, durable revenue commitment rather than a routine vendor arrangement. CoreWeave has been positioning itself as a neutral GPU cloud provider for frontier AI labs, and landing Anthropic alongside its existing roster of customers strengthens that story considerably.

The Broadcom chip deal is harder to size from public disclosures. Broadcom’s announcement grouped Anthropic with Google, so the relative scale of Anthropic’s portion is unclear. What the deal does confirm is that Anthropic is pursuing custom silicon relationships in parallel with its cloud compute strategy. That mirrors the path taken by Google (TPUs) and Amazon (Trainium) and suggests Anthropic is trying to reduce long-run dependence on Nvidia hardware, though no specific chip specs or volume commitments have been publicly detailed.

Together, the two deals point to an Anthropic that is actively locking in infrastructure capacity at multiple layers of the stack, compute, cloud, and silicon, rather than relying on a single vendor. For a company reportedly eyeing a Q4 IPO, that diversification reduces a key risk factor prospective public investors would scrutinize.

Is Michael Burry’s Palantir thesis credible?

Burry’s argument, as reported by Business Insider, is that Anthropic is “eating Palantir’s lunch” by making AI integration easier for enterprise customers. The framing matters because Palantir has long marketed itself as the sophisticated, hard-to-replicate AI platform for government and large enterprise. If Claude-based tooling can undercut that positioning on ease of deployment, Palantir’s premium valuation becomes harder to justify.

Burry has a track record of contrarian calls that eventually prove correct, though the timing on his trades is notoriously unpredictable. His view here is consistent with a broader pattern: several enterprise software vendors have seen deal cycles slow as customers experiment with direct API access to frontier models rather than buying a wrapped platform on top. Whether Anthropic specifically is the beneficiary, versus OpenAI or even open-source models, is an open question the available sources do not resolve.

What is notable is that a prominent short-seller is now treating Anthropic as the reference competitor when making a bearish case against an established public AI company. That kind of named competitive positioning typically follows a period where a private company has reached sufficient market presence to move public-market sentiment. It is a soft but real indicator of how Anthropic’s commercial profile has shifted over the past year.

What does this mean for Anthropic’s IPO trajectory?

None of this week’s news changes the fundamental IPO timeline reported previously, a potential Q4 2026 window. But the deal activity is consistent with a company that is building the infrastructure and partner relationships needed to support the revenue scale a public offering would require.

The CoreWeave and Broadcom agreements add to existing ties with Amazon Web Services and Google Cloud. Having multiple hyperscalers and infrastructure providers in the mix gives Anthropic negotiating leverage and operational redundancy. Both are attributes that public-market investors in capital-intensive AI infrastructure tend to reward.

The Burry commentary, while not a direct Anthropic financial disclosure, adds an unusual dimension: a market figure with credibility among institutional investors is now actively using Anthropic’s competitive rise as the basis for a public short thesis against a rival. That kind of third-party competitive validation, even from an adversarial angle, can influence how buy-side analysts frame Anthropic’s addressable market ahead of any prospectus.

Nothing here constitutes investment advice. Anthropic remains private, and the deal financial terms disclosed publicly this week are limited.