Skip to content
All
News
Sport
Culture
Lifestyle
Related: Anthropic tells people to 'keep thinking'
Anthropic has warned in updated IPO filings that its artificial intelligence systems could imminently achieve recursive self-improvement.
The company stated it might not realise this milestone has occurred until after the fact, as it relies on lagging indicators to track development pace.
Anthropic cautioned that if model capabilities outpace safety, alignment and security measures, it could fail to mitigate catastrophic risks to humanity.
The firm committed to maintaining human supervision over frontier model development, including delaying releases of powerful models like Mythos.
Investors were advised that these precautionary safety measures could increase operational costs and leave Anthropic at a competitive disadvantage.
In full Anthropic says it could soon cross a huge AI threshold – and we might not know about it
Thank you for registering
Please refresh the page or navigate to another…