PfMS 2025 Q&A: Question 1

by | Jun 6, 2025 | Portfolio Management, Portfolio Management Summit 2025

(AI): How are portfolio managers using AI to assist in portfolio prioritization activities? Particularly as it relates to executive visibility & giving executives additional insights into tradeoff conversations.

Synthesized by ChatGPT then edited and enhanced by Te Wu.
Specific contributors include:

  • Cédric Kahl
  • Cristina Niculescu
  • David Vincenti
  • Kris Sprague
  • Lars Hansen
  • Todd Generotzke
  • Anonymous

Artificial intelligence (AI) is steadily reshaping the landscape of project portfolio management (PfM), offering powerful tools to assist professionals in navigating increasing complexity and accelerating decision cycles. At its core, AI’s promise lies in its ability to deliver executive-level insights, support scenario planning, and enhance prioritization processes through predictive and prescriptive analytics. However, despite the growing adoption of these technologies, a recurring theme persists: the gap between AI’s theoretical potential and its practical application remains significant.

Researchers and practitioners alike have noted that although algorithmic approaches have been studied extensively over the decades, their uptake in real-world portfolio decision-making has been limited. This hesitance often stems from perceptions that AI-driven models are opaque, overly complex, and incompatible with the power- or opinion-based decision styles prevalent in many organizations. Moreover, recent scholarship highlights cultural and psychological barriers, with stakeholders expressing discomfort with AI’s lack of transparency and human judgment.

Nonetheless, successful integration of AI into PfM practices is gaining traction in several domains. AI is increasingly leveraged to power scoring models that evaluate project alignment to strategic objectives, value contribution, risk exposure, and resource demand. These models enable dynamic prioritization that evolves with changing organizational contexts and data inputs. Through integration with executive dashboards, AI can deliver real-time, actionable insights—for example, projecting the return on investment when reallocating resources or flagging high-risk projects due to impending bottlenecks.

Key applications include:

  • Scenario Planning: AI generates multiple “what-if” simulations to explore tradeoffs between cost, schedule, and scope under varying assumptions, aiding in contingency planning.
  • Predictive Analytics: By detecting patterns in historical data, AI can forecast potential delays or resource shortages, promoting proactive mitigation.
  • Strategic Alignment Evaluation: Projects can be scored systematically against criteria such as business value and feasibility, supporting more defensible and aligned portfolio decisions.
  • Enhanced Executive Insight: Sophisticated visualizations help surface risks and opportunities, enabling more informed and agile decision-making.

Still, AI does not replace human oversight. It is a tool that augments—rather than substitutes—strategic thinking and judgment. Effective use of AI requires a balanced approach, ensuring that its outputs are comprehensible, validated, and integrated into existing governance frameworks. Portfolio managers must act as interpreters of AI-generated insights, contextualizing recommendations within organizational realities.

Furthermore, the current state of software integration reveals a fragmented picture. While some portfolio tools embed AI to model asset lifecycle interventions and optimize cost trajectories, many organizations still rely on traditional tools like PowerPoint for decision-making communication, with AI insights operating at the margins rather than the core.

The road ahead demands a cultural shift—one that embraces transparency, builds trust in algorithmic tools, and fosters collaboration between technology and leadership. Only then can organizations fully capitalize on AI’s ability to elevate portfolio management from a reactive function to a strategic powerhouse.

Contribution by: Cédric Kahl

Beyond prioritization, AI capabilities should be considered for:

  • Value maximization: Optimizing project mix under constraints (resources, funding, and risk appetite) using prescriptive analytics.
  • Predictive insights: Identifying internal and external risks and opportunities before they materialize, supporting proactive decisions.
  • Scenario planning: Simulating future states and assessing how different strategies perform under uncertainty, improving long-term resilience.

 

Contribution by: Cristina Niculescu

Here are some scenarios we can use in prioritization of portfolio programs:

  • Use AI to give us some criteria of prioritization between projects, based on company strategy and objectives.
  • Give as input the list of projects/programs and their deadlines, and ask for the best delivery plan
  • Use predictive analytics for risk management
  • Use NLP for summarizing project updates, or summary meetings
  • Use AI-powered scenario planning tools like Planview AI or Smartsheet Copilot

Important note: AI helps us but doesn’t replace our judgement and strategic decision-making.

 

Contribution by: David Vincenti

PfMs, like all professionals, can use the standard generative AI tools to create graphics and improve their slides and reports. Additionally, they can use the tools to rapidly generate multiple presentations of data sets for presentation to decision-makers (eg, ROI, IRR, 5-year NPV, 10-year NPV). The caution, of course, is that AI is likely to make errors that are hard to catch (eg, discount rates for more speculative ventures).

 

Contribution by: Kris Sprague

Portfolio managers are increasingly leveraging artificial intelligence (AI) to enhance portfolio prioritization activities, particularly in providing executive visibility and insights into tradeoff conversations. AI offers sophisticated tools and methodologies that help manage the complexity inherent in portfolio prioritization by considering multiple factors and variables. This technology aids in creating scenarios that can mitigate negative impacts on the organization, driven by assumptions and constraints that frame decision-making processes.

  • Scenario Creation: AI can generate various scenarios based on different assumptions, such as allowing costs to vary to maintain contractual completion dates or assuming that contracts can be renegotiated to discount the impact of delays. These scenarios provide executives with a clearer understanding of potential outcomes and tradeoffs.
  • Resource Management: AI can assume elastic talent resources, which helps in addressing resource shortfalls and provides insights into how resource allocation can affect project timelines and outcomes.
  • Strategic Alignment: AI tools can systematically evaluate projects against strategic alignment, contribution to business value, and implementation complexity, ensuring that prioritization decisions are aligned with organizational goals.
  • Enhanced Decision-Making: By providing a robust agent-based model, AI can answer critical questions about the strategic value, feasibility, and overall impact of projects, offering executives a comprehensive view of the portfolio.
  • Executive Insights: AI enhances executive visibility by offering detailed analytics and visualizations that highlight key tradeoffs, risks, and opportunities, enabling informed decision-making and strategic discussions.

In summary, AI assists portfolio managers by offering advanced capabilities to handle complex prioritization challenges, providing executives with valuable insights into tradeoffs and strategic decisions.

This technology supports a more dynamic and informed approach to managing project portfolios, ensuring alignment with organizational objectives and optimizing resource allocation. By integrating AI into portfolio management, organizations can achieve greater efficiency, transparency, and strategic alignment in their project prioritization efforts.

 

Contribution by: Lars Hansen

As a researcher, I’m eager to see what will happen here. Over the past three years, I’ve studied the academic literature on project portfolio management, including a review of 668 publications spanning seven decades of research (Hansen and Svejvig, 2022). One recurring pattern is particularly striking: although algorithmic approaches often appear promising in theory, they are rarely adopted in practice (Hansen and Svejvig, 2023).

This reluctance is often due to perceptions that such approaches are too complex or opaque (“assisted”) and that they conflict with the prevailing decision-making style of organizations. For example, organizations may rely on power-based or opinion-driven decision styles, as suggested by Kester et al. (2011) , which do not align well with data-driven AI methods. Similarly, as explained in the seminal work by Cooper et al. (1999), a data-driven approach may not fit the style of senior management.

So, I’m very curious to see whether organizations will continue to follow historical patterns or if something different will emerge this time. However, I remain skeptical about the role of AI as an autonomous agent in portfolio-level decision-making. Recent research by De Freitas (2025) offers compelling reasons for this skepticism, showing that people often resist using AI because they perceive it as too opaque, emotionless, inflexible, and black box, and, fundamentally, because organizations prefer human interaction.

References:

  • DE FREITAS, J. 2025. Why People Resist Embracing AI. Harvard Business Review, 103, 52-56.
  • HANSEN, L. K. & SVEJVIG, P. 2022. Seven Decades of Project Portfolio Management Research (1950–2019) and Perspectives for the Future. Project Management Journal, 1-18.
  • HANSEN, L. K. & SVEJVIG, P. 2023. Principles in Project Portfolio Management: Building Upon What We Know to Prepare for the Future. Project Management Journal, 1-22.
  • KESTER, L., GRIFFIN, A., HULTINK, E. J. & LAUCHE, K. 2011. Exploring portfolio decision‐making processes. Journal of Product Innovation Management, 28, 641-661.

 

Contribution by: Todd Generotzke

The two areas that come to mind for me are portfolio prioritization scoring models and executive dashboards.  

With the amount of project data available, using AI can help portfolio managers create dynamic prioritization models that can assess projects across multiple metrics including strategic alignment, value, risk, resource demand, time-to-impact, etc. Using AI can help refine these models over time and help identify hidden patterns and weighting criteria based on historical outcomes. The analysis provides objective, defensible rankings of portfolio items, which can serve as powerful inputs for executive review.

AI can integrate with executive dashboards to go beyond traditional status updates and offer predictive insights.  Examples include projects at risk of delay due to resource bottleneck or the reallocation of resources from X to Y can result in ROI increases by Z%.  Insights like this help executives quickly understand trade-offs, consequences, and assist in making faster and better-informed decisions.

 

Contribution by: Anonymous

Portfolio software may contain an AI model that can be added to the investment to help predict intervention points within the asset lifecycle. Decisions can be made for the lowest lifecycle cost where maintenance and capital replacement are optimized. My experience with Scenario Planning is through a manual effort or by optimizing value through software which relies on logic. Presentations are still delivered through PowerPoint and decisions are currently not being influenced by AI.