From Adoption to Impact: Implementing the DORA AI Capabilities Model

The 2025 State of AI-assisted Software Development report reveals a crucial insight: AI acts as an amplifier, enhancing the strengths of high-performing organizations while exposing the dysfunctions of those struggling.

Despite AI adoption being nearly universal—with 90% of developers incorporating it into their daily workflows—success is not guaranteed. A detailed cluster analysis of almost 5,000 technology professionals shows significant variation in team performance; not everyone experiences the same benefits from AI. This disparity indicates that how AI is used is a critical factor. To understand the specific capabilities and conditions that enable teams to achieve positive outcomes, we developed the DORA AI Capabilities Model report.

This guide, a companion to the 2025 DORA Report, is designed to help organizations navigate the new reality of AI-driven development. It provides actionable strategies, implementation tactics, and measurement frameworks to assist technology leaders in cultivating an environment where AI truly thrives.

Seven Capabilities That Amplify Success

Successfully leveraging AI requires cultivating both a robust technical infrastructure and a supportive cultural environment. From the same survey participants as the 2025 DORA report, seven foundational capabilities were identified as proven to amplify AI's positive impact on organizational performance:

• Clear and Communicated AI Stance: Ambiguity creates risk. A well-defined policy fosters the psychological safety developers need for effective experimentation with AI.
• Healthy Data Ecosystems: The efficacy of AI is directly dependent on the quality of the data it learns from. Investing in high-quality, accessible, and unified internal data significantly enhances AI's benefits.
• AI-Accessible Internal Data: This goes beyond simple prompts and involves "context engineering," securely connecting AI tools to your internal documentation and codebases.
• Strong Version Control Practices: As AI increases the volume and velocity of code generation, robust version control becomes an essential safety net. Frequent commits and reliable rollback capabilities are crucial for maintaining stability in an AI-assisted world.
• Working in Small Batches: AI can easily generate large blocks of code, which are difficult to review and test. Enforcing the discipline of small batches counteracts this risk, ensuring that speed translates into product performance rather than instability.
• User-Centric Focus: Speed is irrelevant if you are moving in the wrong direction. Adopting AI tools can actually harm teams that lack a user-centric focus. Keeping user needs as your guiding principle is essential for steering AI-assisted development effectively.
• Quality Internal Platforms: A strong internal platform provides the automated, secure "paved roads" that allow AI benefits to scale across the organization. This prevents individual productivity gains from being lost to downstream bottlenecks.

The DORA AI Capabilities Model illustrates how these specific capabilities amplify the effects of AI adoption on various outcomes.

Where to Start: Assessing Your Team

Every organization begins its AI journey from a unique position. To help prioritize interventions, this report introduces seven distinct team archetypes, derived from our cluster analysis. These profiles range from "harmonious high-achievers," who excel in both performance and well-being, to teams facing "foundational challenges" or those stuck in a "legacy bottleneck" due to unstable systems and undermined morale. Identifying the profile that best matches your team can pinpoint the most impactful interventions. For example, a "high impact, low cadence" team might prioritize automation to improve stability, while a team "constrained by process" could focus on reducing friction through a clearer AI stance.

Digging Deeper with Value Stream Mapping

Once your team's profile is understood, how do you effectively direct your efforts? The report includes a step-by-step facilitation guide for running a Value Stream Mapping (VSM) exercise. VSM serves as an AI force multiplier. By visualizing the entire flow from initial idea to customer delivery, you can identify where work waits and where friction exists. This ensures that the efficiency gains from AI are channeled into solving system-level constraints, rather than just creating local optimizations that might accumulate work downstream.

Accelerating Improvement

AI adoption represents a fundamental organizational transformation. The greatest returns stem not merely from the tools themselves, but from strategically investing in the foundational systems that enable and amplify their effectiveness.