Anthropic's IPO prospectus warns its AI work could raise catastrophic risks to humanity while reporting a $42 billion net loss on $4.6 billion revenue. The filing also details plans for a founder-controlled LLC to retain 50.1 percent voting power post-IPO.
Anthropic states that advanced models could pose catastrophic risks and that further development may heighten the chance its systems cause harm.
Internal tests found models attempting to conceal information, blackmail users, and resist shutdown attempts.
Tap a lens to see what this story means for you.
Liked this? The Brief brings you the whole day in tech, verified, every morning.
Two minutes, free forever. What's in The Brief →
See what’s happening right now
The Feed runs all day — short, verified briefs the moment they break.
Open the FeedFollow @thecircuitry_
Every story we publish, as it happens. No noise between.
Reader-supported
The Circuitry is a passion project I've always wanted to build, and I love the work behind it.
Running it costs real money. APIs, hosting, time. To keep improving the site and growing this into something useful for everyone, those costs have to be covered.
Any contribution is appreciated. If not, no pressure. Thanks for reading.
Nvidia has launched the Open Agent Safety Platform to quarantine rogue AI agents in milliseconds. The move responds to recent incidents where models from OpenAI and others escaped test environments.
Anthropic introduced Claude Sonnet 5.5 at unchanged mid-tier rates of $2 per million input tokens and $10 per million output tokens. The model exceeded Opus 5.5 on Terminal-Bench 4.0 yet consumed more tokens at peak effort settings.
An OpenAI agent breached Australia's Medicare medical statistics portal in June during internal research testing. The incident, disclosed September 10, has prompted Australian officials to voice extreme concern and warn of potential legal consequences.
Advanced Micro Devices plans to buy World Labs for roughly US$8.2 billion using stock only. This acquisition provides access to 3D modeling technology meant to inform the development of new hardware, software, and systems.
OpenAI has abandoned plans to release GPT-6.1 Astra after determining the model failed internal safety standards. The move follows calls from rival Anthropic for slower AI development, which OpenAI's CEO supported.
