Alibaba Bans Claude, Betting Markets Dim on Anthropic IPO

Two pieces of news round out a busy week for Anthropic: a significant corporate ban in China and a fresh read from prediction markets that suggests the company’s IPO timeline may be less certain than the headlines imply.

Key points

• Alibaba has placed Anthropic’s Claude Code on a high-risk software list, banning employees from using it following a “distillation attack” accusation.
• Prediction market traders on Polymarket are not betting on Anthropic as the first major AI IPO in 2026, despite widespread media coverage of the company.
• The crowd favorite on those markets is reportedly Databricks, with OpenAI and Anthropic trailing, according to the same report.
• Anthropic’s 20-year TeraWulf data center lease, reported yesterday, continues to signal long-term infrastructure commitment regardless of public market timing.

What does the Alibaba ban actually mean?

The details from CNBC are specific: Alibaba placed Claude Code, Anthropic’s coding assistant product, on an internal high-risk software list. The stated reason connects to accusations of “distillation attacks,” a term referring to the practice of using one AI model’s outputs to train or improve a competing model.

The ban applies to Alibaba employees, meaning it is a corporate policy decision rather than a regulatory or government action. China is a market Anthropic does not officially serve directly, given existing U.S. export restrictions and the company’s focus on Western enterprise customers. Still, the accusation carries reputational weight. If the distillation framing gains traction, it could complicate Anthropic’s relationships with other large corporate clients who may worry about similar exposure.

No response from Anthropic on the accusation was included in the sourced reporting. The specifics of what triggered the “distillation attack” label remain unclear from available reporting.

What are prediction markets saying about the IPO race?

The Polymarket data, as summarized by Yahoo Finance, flips the conventional read on the AI IPO queue. Databricks, not OpenAI or Anthropic, is the current crowd favorite for which company goes public first. Anthropic, despite being the subject of near-constant IPO speculation, is not leading that market.

This is worth watching for a simple reason: prediction markets aggregate the views of people with money on the line, which can produce a cleaner signal than analyst commentary or media narratives. The market’s lean toward Databricks may reflect several factors. Databricks has been profitable on an operating basis for longer, has a more established enterprise sales motion, and faces less regulatory scrutiny than the frontier AI labs. Anthropic, by contrast, is still burning significant capital on model development and infrastructure, as the TeraWulf lease and other recent deals illustrate.

None of this means Anthropic will not IPO in 2026 or 2027. It means the betting community currently assigns that outcome a lower probability than the volume of Anthropic IPO coverage might suggest.

Does the infrastructure build contradict the IPO hesitation?

On the surface, a 20-year data center lease and a potential Samsung chip deal (reported earlier this week) look like the moves of a company planning to stay private and scale for years. Long-term infrastructure commitments can actually complicate an IPO, since they lock in future capital obligations that public market investors must underwrite.

At the same time, large cloud and compute contracts are often prerequisites for the revenue visibility that IPO underwriters demand. The two dynamics are not mutually exclusive, but they do suggest Anthropic is prioritizing operational scale over near-term market access, which aligns with the prediction market skepticism about a rapid public listing.

Today is relatively quiet on net-new Anthropic developments. The Alibaba ban and the prediction market framing are the substantive additions to the week’s coverage, and both point in a similar direction: Anthropic faces friction in China and less IPO urgency in the market’s eyes than its fundraising profile alone would suggest.