AI in Project Management: What Pain Points It Solves

Advertisement

Reporting burden, media discontinuities, delayed risk signals: What AI and automated document processing take off project managers' plates today—and what they don't.

AI in project management rarely fails because the technology doesn’t work. Rather, it fails because no one defines the problem it is supposed to solve. Project managers spend a large portion of their time on work that does not advance the project: compiling status reports, writing meeting minutes, and transferring figures from quotes, change orders, and invoices into the project plan and budget. At the same time, the demands for oversight and coordination are increasing.

The major standard-setting bodies in the profession responded to this in 2025 and 2026. In June 2026, the Project Management Institute (PMI) published its own standard for AI in portfolio, program, and project management; PeopleCert released a practical guide to AI for PRINCE2; and the GPM German Society for Project Management published a study on changing roles. Their joint conclusion: AI takes over data collection and design work, while verification and decision-making remain the responsibility of humans. This article highlights the specific pain points this affects.

Complexity is the norm, not the exception

At Pulse of the Profession 2026 According to the PMI survey, which polled 2,023 project professionals and 511 executives from 35 countries, 81 percent say projects have become more complex in recent years. 97 percent led at least one complex project in the past year. About one-third of complex projects (31 percent) do not fully achieve their originally planned benefits. Nearly half of the respondents (48 percent) cite faster technology and tool cycles as a driver.

It is noteworthy how PMI describes the consequences: escalations take longer, approvals get stuck at hidden levels, and „Who is responsible for this?“ becomes a recurring question. Among the negative consequences of complex projects, the report cites delayed decisions by stakeholders, at 34 percent.

The German perspective complements the GPM Study: „Project Management in the Age of Artificial Intelligence“ with 300 valid responses from project specialists with an average of just over 18 years of professional experience. Respondents rated increasing pressure to perform—that is, more work in less time—at an average of 4.5. The study cites fragmented project information and insufficiently structured data as obstacles to AI adoption.

Pain Point 1: Reports and logs take up management time

Status reports, meeting minutes, and decision papers follow set formats and are based on information that is already available. This is exactly where project professionals most often use AI: After Latest PMI data from October 2026 77 percent use it for research, analysis, and information gathering, and 69 percent for content, documentation, and project results. The GPM study explicitly lists the creation of minutes, the summarization of status information, and the drafting of reports among the tasks that AI can take on.

The value lies in the draft, not in the approval. An AI can draft a report based on ticket statuses, milestones, and meeting notes. Whether a status indicator is yellow or red, whether a deviation is escalated, and how it is communicated to the client remains the project manager’s decision.

Pain Point 2: Discontinuities between project documents and the project plan

A significant portion of project management relies on documents: quotes and addenda determine the budget, delivery slips and acceptance reports determine the degree of completion, and invoices determine the actual cost status. If these values are entered manually, delays and data entry errors occur. Discrepancies then only become apparent in the monthly closing rather than during ongoing monitoring.

Automated document processing addresses this media discontinuity. According to the Federal Statistical Office, text mining—the automated analysis of texts—is already the most widespread AI application in German companies: 52 percent of companies that use AI are implementing it (reporting year 2025). 27 percent are using AI to automate workflows or decision-making.

It is not the machine alone that reduces errors in the process, but rather the verification architecture. Fields with low recognition reliability are routed to a human; approved values are automatically transferred to the cost plan and status report without having to be retyped; and every correction remains traceable. Vendors such as Konfuzio, on whose portal this article appears, see this as a Manual inspection and post-processing as part of a human-in-the-loop approach. Transmission errors are eliminated, and AI reading errors become apparent rather than slipping unnoticed into the report.

A clean database is a prerequisite. The PRINCE2 AI Practice Guide PeopleCert therefore recommends starting with the data rather than the technology. Among German companies that have considered AI but have not yet implemented it, the following mention 44 percent reported difficulties with data availability or quality and 45 percent cited a lack of compatibility with existing systems.

Pain Point 3: Risks Become Apparent Too Late

In complex projects, risks rarely arise within a single work package but rather at the interfaces: a delayed supplier, a change order that ties up capacity, or a milestone on which three others depend. AI can identify patterns and dependencies across schedule, resource, and cost data that would otherwise be lost in individual reports. According to PMI, 53 percent of project professionals already use AI to identify risks, opportunities, and emerging issues. The GPM study cites early risk detection, scenario planning, and capacity conflicts as typical areas of application.

However, a risk signal is not the same as a risk assessment. Whether a detected dependency is critical depends on context that the data does not include: contractual leeway, political priorities, and a team’s resilience. The new PMI Standard for AI It therefore describes human-in-the-loop practices that teams use to review AI results, determine whether escalation is necessary, and decide when to accept or reject a recommendation.

Pain Point 4: Relief That Turns Into an Increased Workload

AI saves time, but the time saved is quickly filled. The GPM study explicitly warns that AI can both reduce the workload and intensify it at the same time, and recommends incorporating the well-being of project teams into the design process. This aligns with the PMI’s finding that the stress from one complex project carries over into the next.

From a management perspective, this means that anyone implementing AI should first determine how the freed-up capacity will be used—for example, for stakeholder coordination, scenario planning, or quality assurance. Otherwise, the number of projects per project manager will simply increase.

Where AI Reaches Its Limits: Decision-Making and Responsibility

There is one point on which standards-setters agree: decisions, validation, and responsibility remain with humans. This is how the PRINCE2 AI Practice Guide puts it, and the PMI standard enshrines “human-in-the-loop” as a working principle. This is empirically confirmed by a master’s thesis that GPM 2025 honored with the German Project Management Student Award. In their experiment on project prioritization, the groups that relied exclusively on the AI's results and decisions, which yielded the worst results.

The legal framework also relies on verified results. Since the revision by the Regulation (EU) 2026/1744, effective as of July 27, 2026, Article 4 of the AI Regulation requires providers and operators of AI systems to take measures to promote AI literacy among their staff. They are not required to guarantee a specific level of proficiency. Where a works council exists, it is required under § 90 of the Works Constitution Act (BetrVG) to provide training on the planning of workflows, including the use of artificial intelligence. If a project tool tracks the progress of individual employees, this may also affect the right to co-determination under Section 87(1)(6) of the Works Constitution Act (BetrVG) be affected. Whether this is the case must be determined on a case-by-case basis.

Software alone cannot bridge the skills gap

The biggest hurdle isn't the technology. Among German companies that have considered AI but haven't used it yet, 72 percent knowledge gap, 62 percent uncertainty about legal consequences and 60 percent concerns about data protection (Destatis, reporting year 2025). No tool can eliminate these concerns.

Companies therefore need project managers who can lead digital projects methodically. They must tailor tasks so that AI can handle them. They must define verification steps before an AI result is incorporated into a report, budget, or decision. And they must take responsibility for approving these results to clients, data protection officers, and the works council. The GPM study describes this shift as a move from data collection to judgment: AI and data literacy, critical evaluation of AI results, and sound judgment are in demand. Traditional project management skills do not lose their importance in this context; 59 percent of respondents disagree with this assumption.

For professional development, the study recommends integrating AI into existing project management topics such as scheduling, risk analysis, and stakeholder communication, rather than treating it as a standalone add-on module. This approach is followed by methodologically sound professional development programs such as a Project Manager Training Course, which combines traditional, agile, and hybrid project management based on frameworks such as PRINCE2, the PMBOK Guide, or ISO 21502 with the proven use of AI in day-to-day project work. What matters most here is not so much knowledge of the tools as the ability to translate AI results into robust project management.

Project management is shifting from data collection to decision-making

AI and automated document processing address real pain points: They shorten reporting cycles, eliminate inconsistencies between documents and the project plan, and make risk signals visible earlier. Project work becomes more efficient and less prone to errors when data entry is automated and verification is systematically organized. The project experts surveyed by the GPM expect that, by 2030, AI will be able to take over, on average, a good half of today’s management tasks. The study makes it clear that this does not mean a corresponding number of project managers will become redundant; rather, it is primarily the structure of their work that will change.

For companies, this presents a twofold challenge: building the data foundation for AI and training the people who will review and take responsibility for its results. Those who invest solely in software risk automating errors as well.