Want to implement AI? Focus on organizational culture (www.techradar.com)

AI success is less about buying the latest model or platform and more about changing the organizational culture that must host it. Jason Foster argues that the single biggest predictor of whether AI becomes business-critical or a failed proof-of-concept is culture: siloed teams block data flows, risk-averse decision processes trap pilots in endless trials, and fear of job loss inhibits adoption. He’s seen promising projects stall not because the tech failed but because the people, incentives, and processes around it weren’t ready. Practically, Foster prescribes three cultural shifts—control to curiosity (permission to experiment), hoarding to sharing (cross-functional data and problem ownership), and fear to confidence (transparency, early involvement, safe testing). Implementation requires behavioural diagnosis, visible leadership modeling, and embedding AI into day-to-day workflows via aligned incentives, decision rights, performance metrics, and governance that enables scaling. For the AI/ML community this means investing as much in change management, data plumbing, and organizational design as in models, since without shared data, rapid iteration, and empowered decision-making, even the best models and infrastructure will never reach production value.