Why Most AI Projects Fail—and How to Fix Them

Thales Teixeira, a Professor of Practice at UCSD, has highlighted common mistakes organisations make with AI projects. He argues that many initiatives fail due to poor problem selection and immature technology. His insights aim to help businesses avoid predictable errors in AI adoption. Teixeira explained that companies often tackle minor issues with AI, which limits the potential value. He stressed that successful adoption starts with choosing meaningful business challenges. Organisations must also assess the importance of the problem and the maturity of the AI solution.

He used McDonald’s drive-through chatbot as an example. The system’s immaturity caused disruptions rather than improvements. Inaccurate AI can create new bottlenecks instead of boosting efficiency. Teixeira’s advice focuses on avoiding implementation pitfalls. Businesses must carefully evaluate both the problem and the technology. This approach can prevent costly disruptions and ensure AI delivers real benefits.