In its third annual State of AI report, ICONIQ highlights how AI products are now driving nearly half of revenue at surveyed SaaS companies, signaling a fundamental shift where AI is no longer an experimental addition but the core business driver.
- AI products projected to exceed 50% of revenue by 2027
- Open source AI adoption grows to 40%, challenging frontier lab dominance
- Investment shifts toward vertical AI in regulated sectors and agentic AI capabilities
What happened
ICONIQ released the 2026 State of AI report, based on a Q2 survey of 305 executives from companies building AI-enabled software products. The report’s narrative has evolved from early 2026’s focus on validating AI technology to now proving its financial impact and revenue contribution. AI product revenue share at these companies rose from 32% in 2025 to a projected 42% in 2026, with expectations to cross 53% by 2027, representing a majority of business income.
The report reveals a complex AI ecosystem where companies use an average of 3.3 AI providers. Anthropic overtook OpenAI as the most used provider, while Google remains significant. Open source models see increased adoption, growing from 37% to 40%, surpassing proprietary in-house models. Customization and fine-tuning of models have become standard, shifting competitive advantages to controlling the AI stack rather than just model access.
Why it matters
This significant rise in AI’s revenue contribution indicates that AI has transitioned from a side project or future bet to a core component of software businesses’ strategies, operations, and valuations. Companies leading this transition treat pricing, cost structures, and organizational design as integral product decisions linked directly to AI capability rather than afterthoughts.
Furthermore, the shift toward vertical AI applications in challenging, regulated sectors such as financial services and healthcare reflects a strategic move away from generic horizontal AI solutions. These sectors value domain-specific workflows as durable competitive moats. The report also highlights a notable focus on developing agentic AI capabilities, although current performance in production remains a work in progress.
What to watch next
The growing prominence of open source AI models suggests continued pressure on frontier AI labs and a potential redefinition of control dynamics in the AI stack. Monitoring how companies balance renting API intelligence versus running their own controlled models will be key to understanding future competitive advantages.
Investments are expected to prioritize agentic AI capabilities, putting emphasis on reliability and orchestration infrastructure. Observing advancements in these domains and their actual impact on productivity and human intervention needs will reveal how quickly agentic AI matures from ambitious roadmaps to effective operational tools.