Anthropic is now being priced at a $1.4 trillion implied pre-IPO value in on-chain markets, the highest level yet for the Claude maker before any stock market listing. The latest number is up 40% in 24 days, based on pre-IPO trading data tied to instruments listed on Jupiter. The same data puts Anthropic’s implied value up 1,067% since October 2025. The pre-IPO instruments on Jupiter are backed 1:1 by SPV exposure, so the trades are being used as a live read on where investors think Anthropic could land when it finally goes public. Since the company is still private, that leaves tokenized and on-chain private-market products doing the price discovery before Wall Street gets its clean IPO roadshow moment. Onchain traders price Anthropic like a trillion-dollar AI stock before its IPO The latest estimate comes in the wake of reports that Anthropic’s annual revenues have exploded from $100 million in 2023 to $45 billion now. It is a massive increase for the firm in a single calendar year. Over the past twelve months, Anthropic’s annualized revenue increased by 1,400%. It is one of the reasons pre-IPO investors compare Anthropic to major publicly listed tech companies before common stock offerings are available to ordinary folks. Investors also keep an eye on how Anthropic turns its AI assistant Claude into commercial applications. According to FIS, Claude powers the Financial Crimes AI Agent for compliance teams involved in anti-money-laundering investigations. The application is designed to allow compliance specialists to complete their work in minutes rather than hours. The system extracts evidence from banking platforms and analyses customer account activities against the database of known money laundering schemes and prioritizes suspicious cases for investigation. It places the AI into a highly sensitive financial environment, where speed, transparency, and traceability come first. There is no room for messing around with customers’ personal information. “According to FIS, BMO and Amalgamated Bank will be among the first institutions to deploy the agent, with broader availability planned for H2 2026,” FIS noted. FIS also emphasized that Anthropic’s Applied AI team, as well as forward-deployed engineers, are participating in the project. They assist FIS in designing the platform and building other agents using Claude technology. The product combines the FIS banking platform, transaction data, compliance systems with Claude AI models. “Client data will always stay inside FIS-controlled systems,” said FIS. “In addition to meeting the high regulatory requirements for compliance, FIS selected Claude as the right choice for our Financial Crime AI Agent because it was able to reason through complex investigations accurately, explain its findings, and operate safely within a regulatory workflow,” Jonathan Pelosi, Head of Financial Services at Anthropic, said. OpenAI opens its cyber model to Europe as Anthropic keeps Mythos under review While Anthropic is dealing with its own battle of cyber AI access in Europe, OpenAI has announced on Monday that it will be giving access to GPT-5.5-Cyber, which is a cyber version of its latest AI model, to the EU countries. Unlike GPT-5.5, OpenAI hasn’t given a similar preview access to Anthropic’s latest Mythos model yet. OpenAI said that European companies, governments, cybersecurity agencies, and EU institutions such as the EU AI Office will have access to the model. It will first start with preview access for cybersecurity teams vetted by OpenAI. The update comes one month after Anthropic launched Mythos, creating fears about the cyber attacks on crucial software systems. Thomas Regnier from the European Commission said, “We are pleased with OpenAI’s transparency in their intention to allow access to their new model to the Commission.” Thomas Regnier confirmed that OpenAI and the EU Commission had discussed the matter in previous meetings. Moreover, they will continue the discussion this week. “That way, we can monitor the roll-out of the model and address any security concerns,” he said. However, according to Thomas Regnier, the EU is also in contact with Anthropic; however, the discussions are at a different stage than those with OpenAI. “Four or five” meetings with Anthropic have been held; however, discussions are still at a different stage than the solution presented by OpenAI. George Osborne, OpenAI’s Head of Country, said that, “An AI lab such as our own cannot unilaterally judge whether a system is secure because it is reliant on trusted partners operating together in concert.” George Osborne added that “it’s time for the latest cyber AI technology to be made available to the many defenders of Europe, not just to the select few.” According to him, the EU Cyber Action Plan will partner with European policy makers, institutions, and business organizations for access to tools for defense and public security purposes. The smartest crypto minds already read our newsletter. Want in? Join them .
