Anthropic's Prospectus Highlights AI Risks Amid $100 Billion Funding Drive
Anthropic seeks $100 billion in new capital while warning that advanced AI could pose catastrophic risks, prompting concerns about the sector's financing and broader economic stability.
Anthropic, the developer of the Claude AI system, released a prospectus this week outlining a $100 billion capital raise. The document includes a warning that highly advanced AI systems could present "catastrophic or existential risks to humanity" if misused, weaponised, or given extensive autonomy.
The AI sector is experiencing a rapid expansion, with global spending on data centres and related infrastructure projected to reach about $1 trillion this year. In Australia, PwC analysts estimate data-centre investment could hit $157 billion by 2030 and $342 billion by 2036.
Reserve Bank of Australia assistant governor Brad Jones said the market’s optimism about tech firms has driven earnings expectations for the U.S. IT sector to around 45 percent per year for an extended period, noting the unprecedented nature of the boom.
Alphabet’s shares fell 10 percent despite a 108 percent revenue increase in its AI cloud business, as investors weighed the company’s plan to spend up to $200 billion on infrastructure.
The Bank for International Settlements’ chief economist, Phurichai Rungcharoenkitkul, warned that the financing model for AI—often “circular financing” where AI firms fund each other and then purchase hardware or cloud services from the same investors—creates systemic risk. He likened the current AI surge to past booms that ended in sharp corrections, such as the 19th-century railway mania and the late-1990s dot-com bubble.
Fitch analysts projected that a 35 percent decline in tech share prices over six months could push the U.S. economy into recession and reduce its size by 3.2 percent within a year, as firms scale back AI-related construction projects.
ANZ chief economist Richard Yetsenga argued that, despite short-term volatility, AI infrastructure could deliver long-term productivity gains comparable to historic technological advances.
Australian Signals Directorate director-general Abigail Bradshaw highlighted that AI is reshaping cybersecurity, offering both new defensive capabilities and enabling faster, more precise attacks by malicious actors.
The prospectus underscores the dual nature of AI development: while it holds the promise of addressing major challenges, it also raises profound safety and economic concerns that will require policy responses.