%20(1).avif)
2025 State of AI Report: The Builder’s Playbook
A Practical Roadmap for AI Innovation
1. AI Product Strategy Has Entered the Next Maturity Curve
2. Evolving Monetization Models Reflect AI’s Unique Economics
3. Talent Strategy Is a Differentiator
4. AI Budgets Are Increasing Fast and Showing Up in Real P&L Terms
5. Internal AI Adoption Is Expanding, But Not Evenly
Final Thoughts
AI has entered a new chapter: from hype to hands-on. Creating and scaling AI products is becoming the pivotal arena for competitive edge. Our 2025 State of AI report, The Builder’s Playbook, shifts focus from AI adoption to the how-to of AI execution, unpacking what it takes to conceive, deliver, and scale AI-powered offerings end-to-end.
Drawing on proprietary April 2025 survey results of 300 executives at software companies and in-depth interviews with AI leaders across the ICONIQ community1, this report offers a tactical roadmap for translating generative AI intelligence into a durable business advantage.
Across all dimensions - from infrastructure to GTM and talent - we believe the leading AI builders are defined not just by model sophistication, but by their strategic agility, cost discipline, and willingness to experiment fast.
Below, we break down five key chapters from the report and what they mean for teams actively building with AI.
1. AI Product Strategy Has Entered the Next Maturity Curve
AI-first companies are moving much more quickly to get their products to market compared to those just adding AI to existing offerings. In fact, nearly half (47%) of AI-native companies have reached critical scale and proven market fit, compared with just 13% of companies building AI-enabled products.
What they’re building: Agentic workflows and vertical applications dominate. Nearly 80% of AI-native builders are investing in agentic workflows, or autonomous systems designed to take multi-step actions on behalf of users.
How they’re building: Companies are converging on multi-model architectures to optimize for performance, cost, and use case specificity, with 2.8 models used on average per respondent for customer-facing products.
.png)
2. Evolving AI Pricing Models Reflect Unique Economics
AI is changing how companies set prices for their products and services. According to our survey, many are now using hybrid pricing models, combining a basic subscription with charges based on how much you use. Some companies are even experimenting with pricing completely based on usage or the actual results customers achieve.
While a good number of companies currently include AI features for free, more than one-third (37%) plan to adjust their pricing in the next year, to better reflect the value customers get and how much they’re using the AI features.
.png)
3. Talent Strategy as a Differentiator
AI is not just a technology problem. It is an organizational one. Most top builders are assembling cross-functional teams with AI/ML engineers, data scientists, and AI product managers.
Looking ahead, most organizations expect 20-30% of their engineering team to be focused on AI, with high-growth companies projecting up to 37%. But survey results show that finding the right talent remains a bottleneck. AI/ML engineers take the longest to hire of any AI-specific role, with an average time-to-fill exceeding 70 days.
Sentiment around the pace of hiring is split. While some feel they’re on track, 54% report falling behind, most often due to a limited pool of qualified candidates.
4. AI Budgets Are Increasing Fast and Showing Up in Real P&L Terms
AI-enabled companies are allocating 10-20% of R&D budgets to AI development. That number is growing across every revenue band in 2025. This shift underscores just how central AI has become to product strategy.
As AI products scale, the cost mix often shifts. In the early stages of product development, talent is generally the biggest expense - this includes hiring, training, and upskilling. But as products mature, cloud costs, model inference, and governance start to make up the majority of spend.
5. Internal AI Adoption Is Expanding, But Not Evenly
Internal use of AI is growing fast, but not everywhere equally. Even though most companies surveyed give around 70% of their employees access to internal AI tools, only about half actually use them regularly. Getting employees to adopt AI is particularly tough in larger, more established businesses.
What works? High-adoption organizations, where 50% or more of employees use AI tools, deploy AI across seven or more internal use cases on average. These include coding assistants (used by 77% of respondents), content generation (65%), and documentation search (57%). Productivity gains range from 15 to 30% in these areas.
The AI Tooling Ecosystem Is Fragmented but Maturing
We surveyed hundreds of companies to understand which frameworks, libraries, and platforms are running in production today. The result isn’t a ranking. It’s a real-world snapshot of the tools developers are using across categories.
Below, you’ll find a quick tour of the most widely used tools in alphabetical order. If you’re building something different or doubling down on a rising alternative, we’d love to hear from you: ICONIQGrowthInsights@iconiqcapital.com.
.jpg)
[1] Survey responses include some but not all ICONIQ Venture and Growth portfolio companies as well as companies not part of ICONIQ Venture and Growth’s portfolio. Please refer to the full report for additional information here.