In late September 2026, Anthropic finds itself under pressure from multiple directions: the US Federal Trade Commission has launched a broad investigation into the company and OpenAI, a reported IPO prospectus shows steep losses despite massive investments, and Anthropic reportedly plans to spend $518 billion on cloud capacity and data centers.
- FTC launches broad investigation into Anthropic and OpenAI, including a consumer-protection probe into autonomous AI agents
- Reported IPO prospectus from Anthropic shows steep losses despite major investments, according to Yahoo Finance
- Anthropic reportedly plans to spend $518 billion on cloud and data centers, with over $100 billion already committed to Amazon, according to The Motley Fool
- Claude Sonnet 4.5 is being deprecated with just 61 days' notice
- Investor Michael Burry calls for a market crash to prevent IPOs from Anthropic and OpenAI, according to Benzinga
Why is the FTC investigating Anthropic?
The Washington Post reports that the FTC has launched a broad investigation into Anthropic and OpenAI. According to a report by the SOFX Report, this includes a consumer-protection probe related to so-called 'rogue AI agents' — autonomous AI agents that could behave in uncontrolled ways. The probe also involves the organization METR alongside Anthropic and OpenAI.
How big is Anthropic's IPO and infrastructure bet?
According to Yahoo Finance, a reported IPO prospectus from Anthropic shows that the company continues to post steep losses despite aggressive investments in AI. PitchBook analyzes the market impact of a potential IPO in a report on the third quarter of 2026. At the same time, The Motley Fool reports that Anthropic plans to invest $518 billion in cloud infrastructure and data centers — with more than $100 billion of that already committed to Amazon.
What does the deprecation of Claude Sonnet 4.5 mean?
According to MIXED Reality News, Anthropic has announced the deprecation of Claude Sonnet 4.5 — with 61 days' notice. That's just one day above Anthropic's own minimum notice period for such announcements.
How is the competition responding to Anthropic's market position?
According to a report by the Korean newspaper Chosun Ilbo, Google's Gemini 4 Argon model outperforms both OpenAI and Anthropic in performance tests. The race for technological leadership in large language models remains tight.
| Topic | Key content |
|---|---|
| FTC investigation | Broad investigation into Anthropic and OpenAI, including a consumer-protection probe into autonomous AI agents together with METR |
| IPO preparation | Reported prospectus shows steep losses despite heavy investment; PitchBook analyzes market impact for Q3 2026 |
| Infrastructure investment | $518 billion planned for cloud and data centers, with over $100 billion already committed to Amazon |
| Product cycle | Claude Sonnet 4.5 deprecated with 61 days' notice, just one day above the minimum standard |
Markets should 'tank hard' to stop OpenAI and Anthropic from going public — 'for the benefit of humanity.'
Michael Burry, according to Benzinga
The combination of regulatory pressure, financial losses despite record investments, and critical voices from the investor world shows that Anthropic finds itself at a pivotal stage. Whether the planned billion-dollar investments and a potential IPO can succeed despite FTC investigations and market skepticism should become clear in the coming months.
