Companies have been actively adopting artificial intelligence (AI) for about a year now. During this time, many companies have experimented with AI technology and undertaken various projects, but a significant number of these projects have failed. According to a recent report by RAND Corporation, 80% of AI projects fail, which is more than twice the failure rate of other IT projects.
However, what is crucial here is that these failures are often attributed not to technical issues, but to human factors. In this article, we will explore the reasons behind AI project failures and investigate specific solutions to address them.
1. Five Key Causes of AI Project Failures
RAND Corporation interviewed 65 AI experts and identified five major reasons for AI project failures. The most significant issue cited was misalignment of goals among stakeholders. Many companies fail to clearly understand the problem AI is supposed to solve, or they proceed with unrealistic expectations. This is a common issue among domestic AI startups as well.
- Misalignment of goals among stakeholders: Many companies begin projects with misguided perceptions of what AI can achieve. They fall into the trap of trying to apply the omnipotent image of AI seen in sci-fi movies or media directly to real-world projects.
- Lack of data and infrastructure: To succeed, AI projects require high-quality data and the infrastructure to process it. Many companies proceed with projects without securing these necessary resources, leading to failure.
- Unrealistic expectations: AI is not a technology that can solve every problem. Attempts to apply AI to overly complex or unsolvable problems usually end in failure.
- Misunderstanding of AI technology: Overcomplicating the technical aspects or attempting to adopt the latest AI technologies indiscriminately can lead to project failure. This phenomenon, known as the ‘shiny object syndrome,’ occurs when companies become overly fixated on the latest trends, losing sight of the project’s core objectives.
- Poor communication: Collaboration between multiple departments is critical in AI projects. However, many projects fail due to a lack of communication, causing the project to deviate from its intended direction or miss key goals.
2. Strategies for AI Project Success
Despite the high failure rate of AI projects, there is no need to give up. On the contrary, we can learn valuable lessons from companies that have experienced failure. To lead AI projects to success, not only technical skills but also organizational communication, data, and infrastructure preparation are essential. RAND Corporation has outlined the following strategies for success.
- Setting clear goals: Before starting a project, it is crucial to define the problem that AI will solve and set clear goals. All stakeholders must share the same goals to expect successful outcomes.
- Securing data and infrastructure: The success of AI projects depends heavily on how well data is utilized. Therefore, it is essential to secure data and prepare the necessary infrastructure to process it.
- Adopting realistic solutions: AI cannot solve every problem. Distinguishing between solvable and unsolvable problems is important. Attempts to solve overly complex or unrealistic problems with AI are likely to fail.
- Long-term perspective: AI projects often struggle to deliver short-term results. To increase the chances of success, companies must focus on solving problems with a long-term view, often spanning over a year.
3. The Importance of Leadership
For AI projects to succeed, leadership is just as important as technical aspects. RAND Corporation recommends that corporate leaders focus on problem-solving rather than technology and maintain a patient approach. In particular, AI projects often do not show immediate results, and leaders should guide the project with a long-term perspective rather than rushing to change directions prematurely.
Leaders must resolve communication issues within the team and facilitate collaboration between departments. Additionally, they should set realistic goals and make trustworthy decisions based on an understanding of AI technology.
Conclusion: The Key to AI Project Success Is ‘People’
The success of AI projects depends not on the technology itself, but on the people who handle the technology. If companies establish clear goals, foster smooth communication between teams, and develop a solid understanding of AI technology, they can overcome the high failure rate and lead their AI projects to success. Therefore, companies considering AI adoption should choose an approach that prioritizes people over technology.