- The article points out a major gap in the corporate AI wave: leadership builds AI strategies, engineers are excited to use new tools, but engineering managers (EMs) are the ones who must turn everything into actual operating procedures.
- Author Vignesh Durai recounts a conversation with six EMs at a developer conference. All faced the same issue: the company declared itself “AI-first,” but there were no specific policies, budgets, or guidelines for implementation.
- Engineering managers must answer critical questions themselves: whether AI-generated code should be put into production, who is responsible for review, and how to measure performance when AI writes code quickly but testing time lengthens.
- According to the article, middle management is becoming the “AI strategy translator” between leadership and engineering teams, even though this role was never designed for that responsibility.
- This situation is widespread across industries such as healthcare, finance, and SaaS. Many managers have to build AI governance processes themselves because legal teams or leadership do not have clear answers.
- New arising tasks include writing internal AI usage guidelines, redesigning code review processes for AI-generated pull requests, and providing psychological support for engineers facing career changes.
- Despite a sharp increase in workload, they are still evaluated by old KPIs such as product release speed, employee retention, and sprint stability.
- The author warns of strategic risk when businesses do not support this “strategy translation” layer: each team implements AI differently, leading to inconsistent quality and governance.
- The article also highlights the risk of losing talent when good managers feel their work is expanding but not officially recognized or supported.
- Another risk is legal liability and product quality: if AI-generated code causes a production error, the engineering manager is usually the one to “take the hit,” even though they were never given clear approval authority.
- The author proposes three urgent solutions: establish AI policies before buying tools, officially recognize the “strategy translator” role of EMs, and create a community for sharing AI adoption experiences among managers.
- A real-world example is cited of an EM who established an internal exchange group on AI governance despite having no budget or direction from superiors.
📌 Conclusion: A widespread problem in businesses implementing AI: strategies are announced from the leadership level, but the hardest part—turning AI into safe and efficient technical processes—is pushed down to the technical management level without support. Engineering managers currently handle governance, code review processes, team training, and operational risks simultaneously, while KPIs remain unchanged. If businesses do not formalize this role, AI adoption may become inconsistent, increase quality risks, and lead to the loss of the best managers.
