Enthusiastic supporters (champions) are common for AI initiatives, but they often lack signing authority. The most predictive question for project success is identifying who will actually write the check for the production version, as this person is rarely the champion.
Leadership's expectation of perfection from AI systems is a major red flag. Organizations ready for AI treat inevitable errors as data points for learning and tuning. If a leader would view a 95% accuracy rate as a failure without context, the company culture is not yet prepared for AI deployment.
Instead of abandoning an AI initiative due to a lack of preparation, the correct first step is a focused 'readiness project.' This involves documenting workflows, auditing data, and clearing governance hurdles. This pre-project determines if AI is even the right solution.
The success of an AI project is less about technology and more about a company's existing project management discipline. If a company's past software projects consistently ran over budget, its AI projects will likely follow the same pattern, but with greater variability and cost.