Anthropic has warned investors of the risks associated with the behaviour of state-of-the-art IS models

In its IPO prospectus, Anthropic warned investors that its technologies could pose a “threat to the survival of humanity”, according to the Financial Times.

Almost a third of the voluminous document, circulated amongst a select group of partners, Anthropic devoted to a detailed description of ‘risk factors’, in particular the ability of its increasingly sophisticated models to resort to manipulation, blackmail and other forms of unpredictable behaviour.

The company also informed investors of the extremely high concentration of its client base: last year, nearly a quarter of its revenue came from just two clients, sources familiar with the document told the FT.

All of this makes it difficult for investors to assess one of the leading US companies in the field of artificial intelligence, the newspaper reports. In its IPO prospectus, the developer of Claude also stated that it plans to spend $518 billion over the coming years to meet its commitments regarding cloud services, computing power and infrastructure required to sustain its rapid growth.

During its May funding round, Anthropic was valued at $965 billion, and its shareholders believe it will be able to place shares based on a valuation of over $2 trillion, attributing their optimism solely to the company’s exceptionally high growth rates.

However, concerns about the risks posed by the development of AI could temper this optimism, writes the FT.

Last week, Anthropic’s chairman, Dario Amodei, told a meeting of the UN Security Council that AI could pose a threat to humanity, calling it “the most important global security issue facing the world today”.

Earlier this month, Amodei issued a public call to curb the pace of development of advanced AI models. OpenAI’s chairman, Sam Altman, and Elon Musk have both supported Amodei’s proposal for collaboration in the sector, with the aim of slowing down development and strengthening safety measures.

According to the IPO prospectus, Anthropic recorded an operating loss of more than $8 billion last year against a backdrop of high computing costs. Its annual revenue grew nearly 12-fold to $4.6 billion. The company’s operating expenses last year totalled nearly $13 billion.

In the second quarter of this year, Anthropic generated revenue of $11.5 billion.

- Реклама -