The State of AI Report (Battery Ventures)

Summary

• AI-exposed companies represent 50% of S&P 500 value and added ~$18T in market cap since November 2022, driving ~75% of total index gains.p.5
• Cloud provider run-rate revenue reaccelerated to $285B in Q3'25 as generative AI workloads expanded across enterprise infrastructure.p.6
• Cloud hyperscalers deployed $329B in CapEx over the last 12 months but remain capacity-constrained with $1.2T in revenue backlog.p.7
• Leading foundation model providers achieved $10B in ARR within 2.5 to 3.5 years, pacing 5x faster than prior software platform incumbents.p.10
• The total addressable market for AI software is projected at $4,100 billion, driven by $1,460 billion in human labor displacement and automation.p.18

Key insights

1. The market capitalization of the S&P 500 AI basket increased by 185% from November 2022 through late 2025 compared to 28% for non-AI constituents.p.5
2. Cloud hyperscalers face a 4x demand overhang with $1,205B in remaining purchase obligations against $329B in trailing twelve-month CapEx spend.p.7
3. OpenAI reached $10B ARR in 2.5 years and Anthropic in 3.5 years, compared to 8 years for Google and 10 years for AWS.p.10
4. Application gross margins are projected to expand from 0%-30% up to 80%+ as token costs deflate and pricing shifts toward business outcomes.p.14
5. Public SaaS 1.0 companies lacking an explicit AI growth narrative trade at a ceiling of ~6.0x EV/NTM revenue regardless of underlying growth rate.p.16
6. Target annual recurring revenue growth for AI-native companies ranges between 5x–10x, compared to 2x–3x for SaaS 1.0 benchmarks.p.21