False Confidence in Using AI in Projects
| AI is becoming part of everyday project management. Project teams use it to analyze schedules, identify risk, forecast costs, recommend resource allocation, and summarize project performance. These activities can now be performed in seconds, not hours or days. However, reliance on AI can introduce the risk of overconfidence in the outputs. The danger is believing that AI output is more accurate than it really is. Several factors can make AI less reliable. Data quality. If the data is incomplete, inconsistent, or biased toward dissimilar projects, the recommendations will reflect those weaknesses. Model training. Every project environment is different, and no model can fully account for all aspects of project type or organizational strategy. Probability theory. AI is based on estimating what is most likely to happen. When probabilities are interpreted as guarantees, project professionals can become overly optimistic. Human behavior. When AI consistently produces useful recommendations, project teams may stop questioning the results. Assumptions and decisions go unchallenged because they are system-generated. False confidence gradually replaces critical thinking. Project leaders can reduce the risk of overconfidence through disciplined oversight. AI recommendations need to be continuously validated and monitored. AI offers exceptional decision-support capabilities, but the responsibility for interpreting results remains with the project leader. Organizations that gain the greatest benefit from AI will be those that implement governance structures and understand when to question AI results. |
Posted on: August 24, 2026 08:00 AM
| Permalink |
Comments (1)
How Projects Can Start Strong
| The most significant improvements to the methodology for managing projects can be found in the book How Big Things Get Done by Bent Flyvbjerg and Dan Gardiner. The research findings were mainly targeted for megaprojects, although some aspects surely apply to all projects. One insight is to get off to a good start. This seems intuitive, yet many project plans are rushed or planned for updates as the project progresses. Once funding is in place, it can be difficult to wait before starting a project, especially with stakeholders who expect immediate, visible progress. Being proactive and creating comprehensive, accurate project plans takes time. However, with AI-based solutions, the ability to get off to a good start changes this. A project manager can use AI to validate the project budget, duration, and risks. The project scope can be analyzed for gaps and inconsistencies. If set up properly, AI analysis can take minutes or hours instead of days or weeks. Getting off to a good start means that within the first months of the project everything runs according to plan. Research suggests that a strong start increases the likelihood of overall project success. Maintaining that momentum serves as an incentive for project professionals to continue using AI to predict risks and minimize other potential variances. As always, using AI requires knowledge, data collection, and effective implementation. Although I often promote an overall integrated project solution for applying AI in project management, using AI for individual project documents is a reasonable approach to get your project off to a good start. |
Posted on: August 17, 2026 08:00 AM
| Permalink |
Comments (1)
Top Three Ways AI Reduces Project Waste
| PMI estimates that approximately 11% of project investment is wasted due to inefficiencies. While some inefficiency is unavoidable, much of it results from delayed decisions, poor information, and teams working on activities that add little value. AI can be used to significantly reduce the most common causes. 1. Reduce Time Collecting and Reporting Information. AI can be used to automatically collect data from schedules, project documents, financial systems, collaboration tools, and field sensors to produce real-time dashboards. Once set up, project professionals spend less time chasing data and producing reports. Instead, they focus on interpreting results, resolving issues, and supporting the project team. 2. Detect Problems Before They Become Expensive. Waste can occur because risk events are only detected after they occur. AI can be used to continuously look for signals or patterns that people may overlook. For example, AI may detect declining productivity, an increasing number of design changes, and supplier delays occurring together. The issue is investigated, and corrective action is taken to prevent milestone delays. 3. Improve Resource Allocation. Resources may be assigned on a fixed basis at the start of the project and not adjusted as the project progresses. AI can be used to continuously evaluate workload, resource skills, priorities, resource availability, and project constraints. The output is recommendations for improved resource assignments and task sequencing. The objective is to enable project leaders to select the best option to deliver the greatest value with the least disruption. AI can be used to help project professionals make better decisions sooner. Reducing a portion of the 11% waste represents a significant opportunity for organizations to improve project performance without adjusting the scope, budget, or schedule baselines. |
Posted on: August 10, 2026 08:00 AM
| Permalink |
Comments (0)
How AI Can Inherit Human Bias
| Many people assume that artificial intelligence makes better decisions because it is objective. Algorithms do not have emotions, preferences, or personal agendas. Yet AI can still reproduce many of the same biases that influence human judgment. The bias usually does not exist in the algorithm itself. Instead, bias enters through the training data, the assumptions built into the model, the objectives selected for optimization, and the thresholds that trigger decisions. As a result, AI models can reproduce the same cognitive biases as project managers. One common example is optimism bias, described extensively by Bent Flyvbjerg in project research. Project teams often underestimate costs and schedules while overestimating the likelihood of success. AI models can make the same mistake if they are trained on unusually successful projects or configured with overly optimistic assumptions. Similarly, anchoring bias can persist when early budgets or schedules remain embedded in forecasting models long after new information suggests they should be revised. AI can also reinforce uniqueness bias. Organizations sometimes believe their projects are too different to benefit from historical comparisons. When models rely only on internal data and ignore external benchmarks, they reinforce that same belief. Overconfidence bias presents another risk. Project leaders may trust AI recommendations simply because they appear mathematical and objective. Similarly, the model itself may project a level of statistical confidence that the underlying data cannot justify. Perhaps the greatest risk is that these biases become less visible once they are embedded in an analytical model. Decisions begin to appear objective because they are supported by algorithms, even when the underlying assumptions remain flawed. Effective project governance requires questioning not only the decisions produced by AI, but also the data, assumptions, objectives, and design choices that shape those decisions. AI should strengthen human judgment, not simply automate its biases. |
Posted on: August 03, 2026 08:00 AM
| Permalink |
Comments (4)
Three Common Questions About AI in Project Management
| Since first speaking about AI in project management at PMI events in 2018, I have been asked hundreds of questions by conference attendees, workshop participants, students, and project leaders. Three questions appear more often than any others. 1. How much data is required to implement AI in a project? The answer is that “it depends”. In project management, useful results can often be achieved with relatively small datasets. For example, optimization techniques based on genetic algorithms work effectively with only project-specific data. In other situations, success depends on finding a set of similar historical projects. To validate the budget for mass transit projects, I used just over 100 projects and 16 project characteristics per project, yielding a significant result (p < .001). For a complex megaproject, compiling a list of risks might require more historical data. So, the amount of data required depends on the problem being solved and the AI method being used. 2. How do I control the use of AI within my project? The answer is governance. Organizations should establish a clear governance plan that defines how AI will be used, what data can be accessed, who is accountable for the process, and how outcomes are applied. AI adoption is not simply a technology initiative; it is also a governance and leadership challenge. 3. What proof is there that AI works? AI is not being used to manage entire projects (yet), but it can significantly improve specific activities such as scope definition, risk identification, forecasting, resource optimization, and decision support. Evidence can be found in the growing number of organizations deploying AI-enabled project solutions and in the expanding body of research supporting their effectiveness. I am encouraged by these questions because they demonstrate that project professionals are moving beyond curiosity about AI and focusing on practical implementation. Understanding data, governance, and evidence of effectiveness provides a strong foundation for using AI responsibly and successfully within projects. |
Posted on: July 20, 2026 08:00 AM
| Permalink |
Comments (2)
