Following CEO Dario Amodei's call to slow development of increasingly powerful models, Anthropic preparing for potential $2 trillion valuation. Company confidentially filed with US SEC June setting stage for IPO after US midterms November.
Anagha Jayakumar reports draft prospectus reviewed by Reuters argues Al will transform global economy far more profoundly than industrialisation, electricity, internet.
Read more - Uproar Over Al Technology and Ethics
Estimated $2-trillion target more than double $365 bn May, signalling rapid growth. However losing large sums: revenue grew 12-fold to about $4.6 bn in 2025 but operating loss widened to $8.06 bn from $2.98 bn in 2024. Net loss almost $4.2 bn though roughly $34 bn accounting charge tied to financing instruments that could convert into shares. Anthropic nevertheless ended 2025 with $20.28 bn in cash, cash equivalents and short-term investments.
Warnings: Company which long highlighted emphasis on safety claims advanced Al could present catastrophic or existential risks to humanity. Filing discloses Al models have been used in ways that could lead to self-harm, acts of violence toward others, or adverse outcomes despite safeguards. Anthropic not been named in any lawsuit concerning mental health of users. Notably, not disclosed how much spent on safety research and says returns on safety investments currently unclear.
Read more - Agents Gone Rogue
Benchmark: Proposed listing follows SpaceX blockbuster IPO June which valued company at $17.77 bn. Could set benchmark for how public markets value frontier Al companies, with rival OpenAl also preparing to go public.
For India which uses Claude for coding, warning matters: compute remains central - Anthropic spent $7.33 bn on compute and infrastructure in 2025, more than three times 2024 spending, at least $518 bn obligations over coming decade.
Tags: #Anthropic #AIPO #Alrisks #DarioAmodei #2TrillionValuation #ExistentialRisk #Claude #Compute #TechIPO #ArtificialIntelligence