Community Specialist| Project Management InstituteNewtown Square, PA, United States
AI projects present their own unique challenges and opportunities, from evolving technology and data dependencies to ethical considerations and unclear success criteria. We invite you to deepen your understanding of AI project management with expert Kathleen Walch!
Whether you’re curious about what makes managing AI projects so unique, figuring out how to navigate roadblocks, or wondering how to position yourself to lead AI work, join this conversation with Kathleen to learn more.
Post your questions about managing AI projects in the comments below by 1pm ET 23 July 2025.
About Kathleen Walch:
Kathleen Walch is Director, AI Engagement and Community at Project Management Institute (PMI). In this strategic role, Kathleen is focused on advancing AI-driven innovation in project management education as an expert practitioner and thought leader. Kathleen joins PMI from the Cognilytica acquisition. At Cognilytica, Kathleen co-developed the Cognitive Project Management in AI (CPMAI) methodology that is in use by Fortune 1000 firms and government agencies worldwide to successfully run and manage their AI and advanced data projects.
Date/Time:
23 July 2025 1-2pm ET
How do “Office Hours” work?
-Comment your question below between now and 23 July 1 pm ET.
-If you have multiple questions, please number your questions so that the expert may address each one clearly.
-Please stay on topic. Questions should be related to managing and implementing AI and data projects.
-Kathleen will answer questions directly on this thread during the scheduled time; visit the thread on 23 July to view the responses. Please note that you will need to refresh the page to view the most recent questions and responses during the live session.
-Kathleen will attempt to answer as many questions as possible during the scheduled “Office Hours” but may not be able to answer each question individually. The discussion thread will be closed at the end of the event, and no additional questions will be accepted after the event.
-When commenting, please adhere to the ProjectManagement.com User Guidelines.
-Please feel free to connect with Kathleen Walch here on ProjectManagement.com as well as LinkedIn to expand your professional network.
Kathleen, I'd like to dig deeper with you on these key issues in artificial intelligence project management:
1) What distinguishes a successful AI project from one that is just a fad within an organization?
2) How should project leadership be transformed to guide multidisciplinary teams in AI initiatives, where uncertainty and experimentation are inevitable?
3) How can we integrate principles of ethics, transparency and bias reduction from the initial phase of an AI project without slowing down innovation?
4) What strategies do you recommend for managing resistance to change in organizations that fear AI replacing jobs?
5) How do you envision the evolution of the Project Manager's role in a world where AI not only supports, but also makes decisions?
3- Transparent, ethical, and responsible AI needs to be a core part throughout your entire AI project. The key is to embed these principles into your AI process, rather than treating them as afterthoughts or compliance hurdles. Because you can build the best AI solution, but if users don’t trust it, they won’t use it.
Successful AI projects use structured, repeatable frameworks like the Trustworthy AI Framework that’s part of PMI’s CPMAI methodology for running and managing AI projects, which integrates five reinforcing layers: Ethical, Responsible, Transparent, Governed, and Explainable AI. Following this framework ensures that trust is built into the system from the start and not bolted on later.
Ethics, transparency, and bias reduction are not barriers to innovation. Rather, they are accelerators of sustainable, scalable, and socially responsible AI. When integrated early and intentionally, this reduces risk, builds trust, and unlocks long-term value. Saving Changes...
Kathleen, I'd like to dig deeper with you on these key issues in artificial intelligence project management:
1) What distinguishes a successful AI project from one that is just a fad within an organization?
2) How should project leadership be transformed to guide multidisciplinary teams in AI initiatives, where uncertainty and experimentation are inevitable?
3) How can we integrate principles of ethics, transparency and bias reduction from the initial phase of an AI project without slowing down innovation?
4) What strategies do you recommend for managing resistance to change in organizations that fear AI replacing jobs?
5) How do you envision the evolution of the Project Manager's role in a world where AI not only supports, but also makes decisions?
4- Resistance to AI often stems from fear, uncertainty, and lack of trust, especially when employees worry about job displacement. To manage this effectively, leadership must go beyond surface-level reassurance and embrace a proactive, top-down communication strategy that is clear, consistent, and human-centered.
Lead with Purpose and Transparency
It’s important that leaders clearly articulate why the organization is adopting AI and address what problem(s) we are solving with AI. Be transparent about how and why and when AI can and should be used. This message should be repeated often and reinforced through multiple channels.
Emphasize Augmented Intelligence
It’s essential to communicate that AI is not about replacing humans, but about augmenting human capabilities and helping humans perform tasks better. This means AI will take over repetitive, dangerous, error prone, or data-heavy tasks, freeing people to focus on higher-value, creative, and strategic work. Framing AI as a collaborative partner helps shift the mindset from fear to opportunity.
Make Communication A Two-Way Street
Top-down messaging should be paired with bottom-up listening. Create forums where employees can ask questions, express concerns, and see how their feedback shapes implementation. This builds trust and reduces the perception that AI is being imposed without regard for its impact.
Equip People with AI Skills and Confidence
Up-skilling employees on AI is critical. When employees understand how to work with AI, and how their roles will evolve, they’re more likely to embrace change and start using AI tools on a regular basis. Training and AI literacy programs should be tailored, practical, and tied to real use cases within their teams. Saving Changes...
Kathleen, I'd like to dig deeper with you on these key issues in artificial intelligence project management:
1) What distinguishes a successful AI project from one that is just a fad within an organization?
2) How should project leadership be transformed to guide multidisciplinary teams in AI initiatives, where uncertainty and experimentation are inevitable?
3) How can we integrate principles of ethics, transparency and bias reduction from the initial phase of an AI project without slowing down innovation?
4) What strategies do you recommend for managing resistance to change in organizations that fear AI replacing jobs?
5) How do you envision the evolution of the Project Manager's role in a world where AI not only supports, but also makes decisions?
5- PMs are a critical part of making AI projects a success. AI is not going to replace the Project Manager. Instead, it is redefining the role as more strategic, human-centered, and future-focused. PMs who embrace this AI evolution will be the architects of successful, responsible AI transformation.
As AI becomes more capable of making decisions, PMs will shift from task execution to strategic orchestration, ensuring that AI-driven initiatives are aligned with business goals, ethical standards, and human values. Saving Changes...
What are the key differences between managing AI projects and traditional project management in the construction sector?
What skills or knowledge areas should project managers focus on to successfully lead AI projects?
Golam
Hi Golam - Construction projects are typically linear, milestone-driven, and governed by physical constraints. Success depends on:
• Clear deliverables and timelines
• Stable requirements and scope
• Compliance with safety and regulatory standards
• Coordination of physical resources and labor
Project managers in construction rely on predictive planning, detailed Gantt charts, and well-defined roles. Risk is often tied to logistics, weather, materials, and site conditions.
AI projects on the other hand are data-driven, iterative, and experimental. They require:
• Continuous validation of data quality and relevance
• Iterative model development and testing
• Ethical oversight and bias mitigation
• Cross-functional collaboration across teams including data science, IT, and line of business
Unlike construction, AI projects often begin with uncertain outcomes. Requirements evolve as models learn, and success is measured not just by delivery but by accuracy, fairness, and business impact.
What are the key differences between managing AI projects and traditional project management in the construction sector?
What skills or knowledge areas should project managers focus on to successfully lead AI projects?
Golam
AI projects are different from traditional software development projects as they’re data-driven, iterative, and often exploratory. To lead AI projects successfully, project managers must evolve their skillset to combine core PM strengths with AI fluency, ethical foresight, and adaptive leadership.
Areas for PMs to focus on include:
Data Literacy: Data is the heart of AI. AI projects revolve around data, not just code like traditional software development projects. PMs must understand data quality, governance, bias, and how data impacts model performance.
Iterative Delivery: AI initiatives require agile, experimental iterations. PMs must be comfortable with evolving requirements, continuous testing, and model retraining. When it comes to AI projects it’s important to “Think Big, Start Small, and Iterate Often.”
Trustworthy AI Practices: PMs must integrate ethical, responsible, transparent, governed, and explainable AI principles throughout the entire AI project lifecycle. This includes bias mitigation, stakeholder transparency, and human oversight.
Soft Skills: PMs managing AI projects need to have great communication skills and excel at communicating risks, progress, and decisions to both technical and non-technical stakeholders. They need to have collaboration skills to be able to work cross-functionally with different teams and stakeholders. They also need to have critical thinking skills to ask the right questions, challenge results, evaluate tradeoffs between accuracy and feasibility.
Understanding of best practices methodology:
To lead AI projects effectively, PMs should also be fluent in CPMAI (Cognitive Project Management in AI) methodology specifically designed for AI and ML initiatives. CPMAI blends traditional project management with AI-specific phases. To learn more you can take this Free Introduction to Cognitive Project Management in AI (CPMAI) course available in 11 languages: https://www.pmi.org/shop/p-/elearning/free...-ai-cpmai/el185 Saving Changes...
1) How do you define project success in AI projects where outcomes are often probabilistic and evolving?
2) What are your top 3 tips for project managers transitioning from traditional IT to AI/data-driven projects?
Looking forward to your thoughts!
Hello Pavan, thanks for joining the conversation!
1- Defining success in AI projects requires a shift from traditional project metrics like time, budget, and scope. Because AI outcomes are often probabilistic (you’ll most likely never get 100 accuracy) and evolve over time, success must be viewed through a more dynamic and multidimensional lens.
First, success should be tied to clear business value. Remember phase 1 of CPMAI is business understanding and addressing what business problem you’re solving. Whether that’s improving decision-making, enhancing customer experience, or driving operational efficiency you need to have your business problem clearly defined up front. Even as the model evolves, the value it delivers must remain aligned with organizational goals.
Second, performance should be measured against meaningful benchmarks rather than perfection. AI models generate predictions with varying degrees of confidence, so success needs to be about meeting or exceeding thresholds that are appropriate for the use case.
Third, adaptability is key. A successful AI project includes mechanisms for continuous learning and iterations, monitoring for model drift (over time your models are going to stop performing as expected), and the ability to retrain models as needed.
Fourth, ethical and responsible AI use is non-negotiable. Make sure to adopt a Trustworthy AI framework to ensure transparency, fairness, and accountability are embedded throughout the AI project lifecycle.
Finally, stakeholder adoption is essential. You need to engage with stakeholders throughout the entire project and ensure that if you built it they will use it (Implementation feasibility is a critical aspect of the AI Go/No Go decision chart at the start of each AI project). Even the most accurate model fails if it isn’t trusted or used. Success includes user satisfaction, integration into workflows, and long-term impact.
Remember, success in AI is not just about what the model does. It’s about how it’s built, how it’s used, how it’s trusted, and how it continues to evolve responsibly. Saving Changes...
1) How do you define project success in AI projects where outcomes are often probabilistic and evolving?
2) What are your top 3 tips for project managers transitioning from traditional IT to AI/data-driven projects?
Looking forward to your thoughts!
2-
• Embrace Uncertainty and Iteration
Unlike traditional IT projects with fixed requirements and linear timelines, AI projects are inherently exploratory. Data quality and model performance may evolve throughout the project. Adopt an agile mindset, plan for iteration, and be comfortable with ambiguity. Remember AI is never “set it and forget it” so you need to understand iteration is just part of successful AI projects.
• Build AI Literacy and Learn CPMAI
To be a successful AI project manager, you don’t need to become a data scientist, but you do need to understand how AI systems work, what makes AI project succeed or fail, and how to manage them responsibly. Familiarize yourself with the CPMAI (Cognitive Project Management in AI) methodology. It’s a vendor-neutral methodology designed specifically for managing AI and ML projects and integrates data readiness, model development, and ethical oversight into a structured framework.
• Soft Skills for Cross-Functional Leadership
AI projects bring together diverse teams (data scientists, engineers, business stakeholders). Strong soft skills of communication, collaboration, and critical thinking are essential. You’ll need to translate technical insights into business value, navigate ethical considerations, and align stakeholders around evolving goals. Saving Changes...
With AI evolving so quickly, what strategies do you find most effective for keeping projects on track despite shifting tools and techniques?
Would love to hear your thoughts.
Hi Hakam,
Focus on the Problem, Not the Tool
Start with a clear understanding of the business problem you're solving. Tools and models may change, but the business problem you’re solving for should remain constant. This helps prevent chasing the latest trend or AI tool and keeps the team aligned on outcomes rather than outputs.
Use Agile and Iterative Frameworks
AI projects benefit from iterative, data-centric methodologies like CPMAI (Cognitive Project Management in AI), which are designed to accommodate experimentation, model retraining, and evolving data. This framework allows for structured flexibility so you can iterate without losing momentum.
Invest in Continuous Learning and Cross-Training
Encourage your team to stay current with emerging tools, but also cross-train on foundational principles like data governance, model evaluation, and responsible AI. This builds resilience and reduces dependency on any single platform or vendor.
Ultimately, staying focused on business outcomes, embracing flexible frameworks, and designing for adaptability allows AI projects to thrive even as the tools and techniques continue to evolve. Saving Changes...
How do you define whether use of AI in a project has been successful or not? I know this may depend on context, but perhaps an example or two would help illustrate!
Hi Amanda, thanks for posting your question!
Success in AI projects is about solving the right problem, delivering real business value and positive ROI, and doing so in a way that is responsible, trusted, and sustainable.
It’s important to remember that AI isn’t the right solution for every single problem. The first step in defining success is determining whether AI is appropriate for your particular need. This begins with evaluating the problem through the lens of the Seven Patterns of AI. If you're unfamiliar with these patterns, you can explore them here: https://www.pmi.org/blog/seven-patterns-of-ai
If your project involves tasks like building an AI-enabled chatbot, implementing facial recognition, or deploying predictive maintenance, then you're likely working on a problem that aligns with one or more of these patterns and AI may be the right fit.
Once you've confirmed that AI is the right approach, the next step is to assess whether the solution delivers real business value. Ask yourself: Did the AI solution produce measurable impact aligned with the project’s goals? Is it solving a meaningful business problem? Positive ROI and tangible outcomes are key indicators of success. For additional insights, this article from the PMI AI blog outlines ten common reasons why AI projects fail—and how to avoid them: https://www.pmi.org/blog/why-most-ai-projects-fail
By knowing what problem you're solving and applying AI responsibly and strategically, you can effectively measure how the solution is performing. Ultimately, every organization must define success for itself, but a strong indicator is delivering measurable business value and achieving a positive ROI. Saving Changes...
PMI Team Member
