TLDR: A significant majority of enterprise AI projects, estimated to be as high as 90-95%, are failing to deliver measurable business value. Experts point to issues such as a lack of clear strategy, poor integration with existing workflows, and a disconnect between technical development and user adoption as primary reasons for these setbacks.
Recent reports and studies highlight a concerning trend in the adoption of Artificial Intelligence within enterprises: a vast majority of AI projects are not achieving their intended objectives. Figures from various sources indicate that the failure rate for enterprise AI initiatives can be as high as 90% to 95%. For instance, S&P Global Market Intelligence’s 2025 survey revealed that 42% of companies abandoned most of their AI initiatives this year, a sharp increase from 17% in 2024. Similarly, RAND Corporation’s analysis confirms that over 80% of AI projects fail, double the rate of non-AI technology projects . A study by MIT’s NANDA initiative further claims that 95% of generative AI implementations in enterprises have ‘no measurable impact on P&L’ .
Experts attribute these high failure rates to several critical factors. One prominent issue is the ‘build-it-and-they-will-come’ fallacy, where sophisticated AI models are developed without sufficient user buy-in, change management, or front-line champions. This often leads to advanced systems, such as contact center summarization engines with high accuracy, gathering dust because supervisors lack trust or agents continue manual processes .
Another significant challenge is the scope and expectation surrounding AI. Many projects are too ambitious, with the belief that generative AI will overhaul businesses overnight. However, AI is often seen solely as a technical problem, rather than a user experience (UX) problem. This leads to a focus on achieving high model accuracy, rather than understanding how users will interact with the product and the relative gains it brings compared to current benchmarks . Peter, from In The Pocket, emphasizes that ‘AI is seen solely as a technical problem, whereas currently, it’s more a UX problem than it is a technology problem’ .
Disconnected organizational teams also contribute to the problem. Product teams, infrastructure teams, data teams, and compliance officers often work in silos without shared success metrics or coordinated timelines, creating significant friction . The data required for AI projects is frequently either nonexistent or of poor quality, further hindering success . Gartner predicts that over 40% of agentic AI projects will be canceled by 2027 due to escalating costs, unclear business value, or inadequate risk controls, leading to wasted budgets and reputational damage .
To beat these odds, a shift in approach is necessary. Companies need a clear strategy, focusing on outcomes rather than just tools. Grayson Lafrenz, CEO of Cadre AI, highlights the ‘Ready-Fire-Aim’ mistake, where companies rush into AI without a clear target. He stresses the importance of ‘building an AI first culture where you’re really transparent and you’re communicating to your team how you’re going to leverage AI’ . Successful implementations involve adapting AI tools to an organization’s existing processes and workflows, rather than expecting the AI to snap into place . Frequent and direct interaction with target users is crucial for ideation and validation, ensuring the AI solution genuinely addresses user problems . Additionally, involving specialists for ethical and legal sensitivities, especially with new regulations like the AI Act, is vital .
Also Read:
- Enterprises Grapple with Generative AI Implementation, Paving Way for IBM’s Solutions
- Healthcare AI Navigates High Failure Rates, Epic’s Strategic Innovations, and Critical Workforce Solutions
While the initial query mentioned ‘Turinton is Rewriting Those Odds,’ extensive research did not yield any specific information about a company by that name actively addressing these challenges. The findings instead reflect a broader industry struggle and the collective efforts of various organizations and experts to navigate the complexities of enterprise AI adoption.
