As businesses prioritize technology advancements for 2025, artificial intelligence (AI) and generative AI are at the forefront. Despite 78 percent of companies integrating AI into at least one business area, over 80 percent of AI initiatives fail, doubling the failure rate of traditional IT projects.
The frequent failure of AI projects often stems from unclear objectives, unrealistic ambitions, and misunderstandings about AI’s capabilities. Many organizations embark on AI projects without clearly defining the issues they aim to address, falsely assuming AI to be an advanced automation tool.
This issue is exacerbated when AI specialists lack essential project management skills and when managers handle AI projects with a conventional approach unsuitable for data-driven initiatives. These challenges underscore the need for companies to refine strategies, ensuring alignment among stakeholders to enhance project success.
To harness the full potential of AI, project leaders must clearly identify the problems they wish to tackle. Success can be measured by reductions in cost, time, resource use, or through improved compliance and user satisfaction.
The Essential Role of Data in AI Projects
Unlike traditional software undertakings, AI projects depend heavily on data—specifically, high-quality and relevant datasets. Project managers need to adopt a data-focused approach to understand the role of data in achieving outcomes. Utilizing frameworks tailored to the demands of data-centric projects is crucial for success.
Managers must pose critical questions to determine if AI is the appropriate solution. What problem are we addressing? Can alternative methods be used, or is AI necessary? If implementing AI, what return on investment is expected? Ensuring the project’s feasibility requires robust data availability and technology alignment with project goals.
Crucial considerations include setting key performance indicators (KPIs), understanding the project’s impact metrics, assessing historical data availability, and tackling potential integration challenges with existing systems.
In AI project management, adhering to the principle of “Think big. Start small. Iterate often” is vital. Initially, focus on a small, well-defined problem. Evaluate data accessibility—internal or external—and prepare accordingly. Starting small facilitates manageable iterations, paving the way for larger challenges, combining business acumen with data insights.
The CPMAI methodology promotes long-term AI success by emphasizing iterative, adaptable processes focused on people and methods.
However, it is important not to solely focus on proof of concept. Real-world data is crucial for AI solutions. Hence, developing pilot projects using real-world data and users allows teams to observe performance, gather feedback, and address issues promptly, enabling scalable solutions based on outcomes.
Continuous AI Project Lifecycle
AI projects require ongoing attention; a “set it and forget it” approach sets them up for failure. Continuous evaluation and model retraining are necessary to avoid misaligned ROI and unfavorable results.
AI is transformative, enhancing project effectiveness. In regions like the Middle East and North Africa, 47 percent of project managers leverage AI for risk prediction and performance monitoring, showcasing its significant impact on overall project outcomes.
The role of project managers is elevated by AI, positioning them as strategic partners in AI-driven transformations. Their responsibilities include adopting methodologies tailored to AI’s unique challenges, such as iterative development and real-world applications. Investing in their training and education is crucial for deriving true business value from AI.
Remember, AI is not a substitute for human intelligence but a tool to enhance capabilities, improving speed, precision, and accuracy in tasks. Ultimately, decision-making remains a human endeavor.
For AI to make a lasting positive impact, responsible implementation is imperative. Addressing ethical concerns, ensuring fairness and transparency, and protecting data privacy are essential in building intelligent, yet trustworthy AI systems.
A mindset of continuous learning will ultimately define whether an organization can thrive in the age of AI.
The author is the Director of AI Engagement and Community at the Project Management Institute.