A recent study by NANDA, an AI company affiliated with the Massachusetts Institute of Technology Media Lab, has revealed that a staggering 95% of generative AI projects fail to make a measurable financial impact. The study analyzed over 300 public AI initiatives, surveyed 153 company leaders, and conducted 52 structured interviews with enterprise decision-makers.
The findings suggest that only 5% of gen AI pilots reach production and generate measurable monetary value, with most projects creating little impact on profit and loss metrics. The study also found that sales and marketing departments are most likely to deploy AI, while finance and procurement are the least likely.
One of the main reasons for the low success rate of AI projects is the lack of contextual awareness in AI models. These models struggle to adapt to specific circumstances, change over time, or remember previous inquiries. Additionally, AI tools often don't fit well with real work processes, leading to "brittle workflows" and poor alignment with day-to-day operations.
The study's findings have significant implications for businesses looking to invest in AI. To increase the chances of success, companies should focus on narrow use cases where AI can bring immediate, measurable savings or productivity gains. Strategic partnerships with knowledgeable vendors can also help.
The industries most positively affected by gen AI are media and telecom, professional services, healthcare, and consumer and retail. As AI technology continues to evolve, it's essential for businesses to carefully evaluate the potential benefits and challenges of AI adoption and develop strategies to maximize its impact.
