Agency, not the tool, decides whether AI works for a team
Skill on one axis, agency on the other: four quadrants with four very different outcomes. Give senior developers agency and the tool starts working. Strip it away and even the best reject it.
Written by
-
Benedikt Stemmildt Founder & Co-CEO
Published: 7 May 2026 Updated: 2 min read
A lot of teams are asking right now why AI works for some teams and not others. I’d say a lot of it comes down to one variable in particular: agency.
That word kept coming up at a CTO bootcamp last week. Once execution stops being the constraint, the bottleneck shifts to decisions. Authority is the title. Agency is whether your input actually changes anything. AI extends it for the people who have it, and exposes who never had it.
Skill on one axis, agency on the other
A simple 2x2 helps me think about it. Skill level on one axis, agency on the other. Four quadrants, four very different outcomes.
Senior plus high agency. These developers know what to build and have the freedom to build it. AI becomes a real multiplier, measurable in lead time from idea to first users, rework rate and the use of what ships. The teams celebrating breakthrough productivity tend to live here.
Senior plus low agency. They have the skill to evaluate AI output but no autonomy to act on it. They see the limitations clearly and can’t course-correct, so the gains stay small or disappear. Frustration builds, sometimes turning into outright rejection.
Junior plus high agency. Steep learning curve, real growth potential. They have freedom to experiment but lack the foundation to evaluate outputs. Some grow fast. Others build technical debt they can’t recognize.
Junior plus low agency. The risky one. Modest gains at best, often an illusion of productivity at worst. They follow AI suggestions without understanding them. No skill to evaluate, no autonomy to question. This is where AI produces debt without anyone noticing.
Where the teams I see actually sit
The teams I see celebrating real productivity gains are senior with high agency.
The teams frustrated after six months of adoption are often senior with low agency. They expected magic and got another tool they can’t fully use.
The teams accumulating debt without noticing are junior with low agency, shipping AI-generated code that gets too little careful review.
The gap is organisational design
AI adoption isn’t really about the tool. It’s about the environment. Give senior developers agency and the tool starts working. Strip agency away and even the best developers will reject it.
The gap isn’t intelligence. It’s organisational design. Agency also sits right at the centre of how roles shift when execution becomes instant.
Where does this match what you’re seeing?
Further reading
Written by
-
Benedikt Stemmildt · Founder & Co-CEO
More on Adoption
- Most AI rollouts change the codebase and leave the team untouched In a Multitudes survey, codebase changes made up 85% of what teams changed for AI, and only 13% named a process change. In our experience, that gap is why most rollouts hit the same wall.
- AI makes teams faster and wears down their engineering instincts Agents speed up output while problem decomposition, debugging and code comprehension fade. The answer is to encode the team's engineering knowledge into a harness.
- Working well with coding agents takes months, not hours Experienced engineers resist agentic coding, and most people quit after a couple of hours. The value only shows once you have put in the time to learn the new skills.
Test the constraint in your system
The Factory Readiness Assessment connects this research to your value stream, your data, and the next decision.
Training on this topic: Agentic Engineering for Engineering and Product LeadersWant to check one team first? The self-check: twelve yes or no questions, three minutes, instant result, no account. Take the self-check