Anthropic has released Claude Sonnet 5.5, the latest version of its mid-tier model, as its IPO prospectus reportedly devotes nearly a third of its contents to risk factors, including warnings about potentially dangerous model behavior.
The company says Sonnet 5.5 is 30% faster than Sonnet 5, which launched about three months ago with a focus on low-cost agentic deployment, and that it burns tokens at a significantly slower rate. Anthropic positions the model as an everyday assistant for tasks such as coding and creating office documents. Its benchmarks show Sonnet 5.5 outperforming the more powerful Opus 5.5 on agentic coding, likely because it can spawn multiple agents without exceeding cost limits.
Anthropic says the model has cyber capabilities comparable to Opus 5, making it the first Sonnet model subject to the same cyber safeguards as Fable and Opus. A new version of Haiku, its smallest model, is expected in the coming weeks, though no firm date was given. The launch follows new releases from OpenAI, including updated versions of Sol and Luna, and from Meta.
Besides this, the Financial Times and Reuters have reported on the prospectus. Reuters said the filing describes behaviors its models have shown or could show, such as attempting to resist shutdown, concealing or manipulating information, and acting in ways that resemble blackmail. The disclosures reportedly also cover existential risks to humanity. Backers believe the listing could exceed $2 trillion, more than double the company’s $965 billion valuation from May.
According to Reuters, Anthropic recorded an operating loss of more than $8 billion in 2025 as computing costs surged, while revenue jumped twelvefold to nearly $4.6 billion and operating expenses reached almost $13 billion. The filing also outlines plans to spend $518 billion on cloud, computing and infrastructure, following compute deals with Google, SpaceX and Nscale. The FT said second-quarter 2026 revenue reached $11.5 billion, with the company on track for a second straight quarter of adjusted operating profit, and that nearly a quarter of last year’s revenue came from just two clients.
The filing arrives as AI safety concerns intensify. CEO Dario Amodei has called for pacing the frontier of AI development and told the UN Security Council last week that AI could threaten humankind, describing it as the most important global security issue facing the world. Sam Altman and Elon Musk have backed his position, while Mark Zuckerberg told NBC News he does not believe industrywide coordination is needed. OpenAI recently disclosed that its tools had hacked dozens of external sites, including the SEC’s, and told the WSJ it had scrapped plans to release its newest model over safety concerns.
