AI Sector Rebound: Anthropic Revenue Surges as Coding Tools Drive Adoption
Six months after widespread fears of an AI bubble, the sector has experienced a dramatic turnaround driven by the release of advanced coding tools like Anthropic's Claude Code. Anthropic's annualized revenue surged from $14 billion to $30 billion in two months, making it possibly the fastest-growing business in history. Software developers are adopting AI en masse, reporting productivity gains of up to 4x. Demand for AI infrastructure now outpaces supply, with Anthropic limiting usage during peak hours and OpenAI scrapping its video-generation app to free up computing power. Academic research confirms productivity improvements: a repeat experiment showed developers completing tasks 20% faster with latest AI tools, reversing earlier findings. The percentage of US businesses with paid AI subscriptions has risen from 25% to over 50% in 2025. Goldman Sachs reports companies overspending on AI by orders of magnitude. Anthropic's new model Mythos, not released publicly due to cybersecurity capabilities, has blown away benchmarks. However, skeptics argue the boom may be limited to coding, warning of overinvestment in infrastructure. A recent MIT study found AI can complete 65% of white-collar tasks, up from 50% in 2024, projecting 80-95% by 2029. The debate now hinges on whether most knowledge work can be automated like coding.
Key facts
- Anthropic's annualized revenue rose from $14 billion to $30 billion in two months.
- Claude Code enables autonomous AI agents to complete programming tasks in minutes or hours.
- Developers using latest AI tools completed tasks 20% faster in a repeat experiment.
- Percentage of US businesses with paid AI subscriptions rose from 25% to over 50% in 2025.
- Goldman Sachs found companies overspending on AI by orders of magnitude.
- Anthropic's Mythos model blew away benchmarks but was not publicly released due to cybersecurity risks.
- MIT study: AI can complete 65% of white-collar tasks, up from 50% in 2024.
- Meta announced layoffs of 10% of workforce, citing AI enabling smaller teams.
Entities
Institutions
- Anthropic
- OpenAI
- Goldman Sachs
- Meta
- Microsoft
- Amazon
- CoreWeave
- Micron
- SemiAnalysis
- Institute for Progress
- Model Evaluation & Threat Research
- Epoch AI
- SK Ventures
- MIT
- University of Pennsylvania
- Axios
- The Atlantic
Locations
- United States