Artificial Intelligence

Anthropic, fundraising for a USD 1 trillion valuation under consideration

The Financial Times writes

8 May 2026

ANTHROPIC AI IA INTELLIGENZA ARTIFICIALE ELON MUSK E DARIO AMODEI COLLABORANO PER PORTARE I DATA CENTER NELLO SPAZIO 7146

2' min read

Translated by AI

Versione italiana

2' min read

Translated by AI

Versione italiana

Anthropic is considering raising tens of billions of dollars this summer to finance a major expansion of its computing capacity. Such a move could raise its valuation to almost$1 trillion and thus put it ahead of rival OpenAI. The Financial Times writes this.

Parola di Anthropic

Meanwhile, SpaceX - Elon Musk's aerospace company - said it had signed an agreement to grant Anthropic access to Colossus 1, a huge supercomputer for artificial intelligence, thus bringing together two of the most prominent players in the AI race.

Anthropic plans to use the additional computing power to increase the capacity available to subscribers of its artificial intelligence assistants, Claude Pro and Claude Max, SpaceX explained in a post.

SpaceXAI will provide @AnthropicAI with access to Colossus 1, one of the world's largest and fastest-deployed AI supercomputers, to provide additional capacity for Claude → https://t.co/nfDR9S822L pic.twitter.com/EQAz0S84m2
— xAI (@xai) May 6, 2026

After the debut of Mythos, OpenAI is also launching a new model for cybersecurity teams

OpenAI announced yesterday that GPT-5.5-Cyber, a variant of its latest artificial intelligence model, is being released on a limited preview basis to selected cybersecurity teams, a month after rival Anthropic caught the attention of investors and government officials with Claude Mythos Preview.

The GPT-5.5-Cyber preview is not intended as a significant leap in offensive or defensive cyber capabilities, but is trained to be more permissive in security-related activities, explained OpenAI.

With the cybersecurity-specific version, authorised teams will be able to more easily use the latest OpenAI model for workflows such as vulnerability identification and classification, patch validation, and malware analysis, the company said. The safeguards built into the publicly available version of GPT-5.5 would have made these tasks more complex.

"GPT-5.5-Cyber allows a small number of partners to study advanced workflows, where specialised access modes can make a difference," OpenAI said in the post.

When it launched Mythos last month, Anthropic decided to restrict access to a select group of companies as part of a new cybersecurity initiative called Project Glasswing. Anthropic's CEO Dario Amodei met with senior members of the Trump administration to discuss the model and its potential, even though the company had been blacklisted by the Pentagon only a few weeks earlier.

Federal Reserve Chairman Jerome Powell and Treasury Secretary Scott Bessent had met with the CEOs of major US banks in April to discuss Mythos, while Vice Chairman JD Vance and Bessent held a call with key technology CEOs ahead of the model's release.