By Markus Kopko, PgMP®, PMP®, PMI-CPMAI™
June 11, 2026
Organizations across industries are exploring how to use AI to improve project, program and portfolio management (PPPM). Yet many struggle with the same question: Where do you start, and how do you scale successfully?
Throughout the past seven article series, we have discussed the building blocks of AI integration: governance, human-in-the-loop oversight, prompting frameworks, AI patterns, agentic coordination, readiness assessment, and the implications of the PMI Standard for AI in Portfolio, Program and Project Management. Each topic is important on its own, but successful implementation requires bringing them together into a practical, actionable framework.
On June 24, 2026, I am excited to share the complete framework in IIL’s webcast, 5 Steps to Integrate AI into Your PPM Practices: A Tactical Blueprint. In this article, you’ll get a preview of the framework, the purpose and desired output of each step.
This article is your roadmap. In the webcast, we’ll take the next step—explore practical execution strategies, templates, real-world examples, and take your questions in a live Q&A to help you put the framework into action.
Why Sequence Matters
Most AI adoption in project management follows an opportunistic pattern: someone discovers a tool, tries it on a task, gets a useful result, and tells the team. That approach produces scattered wins. It does not produce sustainable integration. McKinsey’s 2025 global AI survey, “The State of AI, 2025”, found that 88% of organizations use AI in at least one business function, but nearly two-thirds have not scaled beyond pilots or experimentation.
The 5-step framework for integrating AI into PPPM practices is designed to prevent the two most common failure modes. The first common failure is premature tool adoption: buying or deploying AI before the organization has the data quality, process maturity, or governance structures to use it effectively.
The second common failure is governance as afterthought: deploying AI first and adding oversight later, which is how shadow AI, accountability gaps, and unreviewable outputs accumulate.
Five Steps to Integrating AI into PPPM
The five steps addresses the problems of premature tool adoption and governance by establishing the foundation before the deployment.
Step 1: Assess Your Readiness
Before selecting tools, patterns, or use cases, assess where you stand. The PM’s AI Readiness Checklist: 5 Questions Before Your First Integration provides the five-question diagnostic: Is your data consistent? Are your processes defined? Does your team have pattern literacy? Do you have a human-in-the-loop protocol? Does leadership support AI adoption with resources?
The readiness assessment is not a pass/fail gate. It is a prioritization tool. If your data governance scores Ad Hoc on Info-Tech’s readiness scale, that becomes your first action item, not your AI pilot. The PMBOK® Guide, Eighth Edition, includes a dedicated Governance performance domain for this reason: governance structures must exist before the capabilities they govern are deployed.
Output of Step 1: A prioritized list of readiness gaps, ranked by their potential impact on the effectiveness, adoption and scalability of AI within PPPM practices.
Step 2: Select Your Patterns
Once you know where you stand, identify which of the seven AI patterns apply to your highest-value project management problems. The seven CPMAI patterns are Anomaly Detection, Classification, Forecasting, NLP, Optimization, Recommendation, and Computer Vision.
The goal is not to implement all seven AI patterns. It is to match the right pattern to the right problem. Most project managers will start with NLP (drafting, summarization, extraction) and Forecasting (schedule and cost prediction). The CRISP Framework for project management prompts – Context, Role, Instruction, Scope and Parameters) provides the prompting structure to get useful output from NLP-based tools immediately.
Output of Step 2: A prioritized shortlist of two or three AI patterns mapped to specific, recurring project management tasks where they add measurable value through improved efficiency, quality or decision-making.
Step 3: Design Your Governance
This is the step most organizations skip, and where most AI integrations fail. Governance for Project Teams: Why Policies Alone Are Not Enoughestablishes the principle that governance at the team level is where AI adoption succeeds or fails.
AI in Project Management Needs Human Judgment added the requirement that human-in-the-loop is not optional. Governance design for AI in PM contains consists of three key elements: decision boundaries, accountability, and auditability.
Decision boundaries define which outputs can be used directly, which require human review, and which require approval. Accountability establishes who is responsible for the quality of AI-assisted deliverables. Auditability ensures that AI-generated outputs are traced, reviewed, and documented. These three elements extend to AI agents, where autonomous action requires stricter boundary definitions as discussed in AI Agents in Program Management: From Coordination Tool to Decision Partner.
Output of Step 3: A documented governance protocol that clearly defines review requirements, assigns accountability, and establishes audit procedures for each AI use case.
Step 4: Run a Controlled Pilot
With readiness assessed, patterns selected, and governance designed, you are ready to pilot. Not experiment. Pilot. The distinction matters: an experiment tests whether AI works. A pilot tests whether AI works within your governance framework, with your data, on your team’s actual tasks.
A controlled pilot has four characteristics. It targets one specific task (not a workflow). It runs in parallel with the existing process for at least one cycle. It measures output quality against defined criteria, not just speed. And it includes a structured review at the end that evaluates both the AI output and the governance protocol.
MIT’s Project NANDA reported in its 2025 GenAI Divide study that 95% of AI pilots deliver no measurable business impact (MIT Project NANDA, The GenAI Divide, 2025). The primary cause is not technology failure. It is the absence of the readiness, pattern selection, and governance work that Steps 1 through 3 provide. A pilot without that foundation is not a pilot. It is a demo.
Output of Step 4: A documented assessment of pilot results, including output quality metrics, governance performance, lessons learned, and a clear recommendation on whether to proceed with scaling.
Step 5: Scale From Evidence
Scaling means expanding from one task to adjacent tasks, from one project to multiple projects, from one AI pattern to complementary patterns. It does not mean deploying AI across the organization simultaneously.
Scaling decisions should be based on pilot evidence: Where did AI output meet quality thresholds? Where did the governance protocol hold? Where did it need adjustment? The readiness gaps that were not addressed before the pilot will surface here. Address them before expanding scope.
For program and portfolio managers, scaling also means evaluating whether agentic AI applies to cross-project coordination tasks: dependency monitoring, resource allocation optimization, or stakeholder communication synthesis. These applications require the governance framework from Step 3 operating at program level, not just project level.
The principle from AI Agents in Program Management: From Coordination Tool to Autonomous Coordinator applies to scaling as much as to agent deployment: expand based on proven results. not enthusiasm alone.
Finally, the last article in this series covers PMI’s Standard on AI’s 5 themes—governance, human oversight, cross-domain competence, data readiness, and ethics—and what they mean for your career.
Output of Step 5: A structured scaling plan with clearly defined scope, measurable success criteria, governance requirements, and a timeline aligned to organizational readiness milestones.
The Full Framework
This article provides the framework and rationale for integrating AI into PPPM practices. Join me on June 24, 2026 for a deep dive into implementation. We will share practical templates for each step, real project examples, an in-depth look at governance design, and a live Q&A.
As a member of the Core Development Team for the PMI Standard on AI in Project, Program, and Portfolio Management, I designed this framework to align with the evolving direction the profession. The five steps are not a theoretical model. They are a practical roadmap for avoiding the common pitfalls that undermine AI adoption: deploying too quickly, overlooking governance, and running pilots that generate interest but fail to deliver measurable outcomes.
Key Takeaways
- The five steps for AI adoption into project, program and portfolio management practices are assess readiness, select patterns, design governance, run a controlled pilot, and scale from evidence.
- Sequence matters. Deploying AI before establishing data quality, process maturity, and effective governance leads to stalled pilots and tools that never gain traction.
- Each step is explained in the preceding articles in this series: governance (article 1), human oversight (article 2), prompting (article 3), patterns (article 4), agent coordination (article 5), readiness (article 6), and PMI’s Standard on AI (article 7).
- A pilot without readiness assessment, pattern selection, and governance design is not a pilot. It is a demo.
This is the final article in a series leading up to the IIL webcast, 5 Steps to Integrate AI into Your PPM Practices: A Tactical Blueprint, on June 24, 2026. Register at https://www.iil.com/your-ai-advantage-practice-habit-strategy/
Markus Kopko is a strategic project and AI transformation expert with over 25 years of experience in project, program, and portfolio management. He contributes to the Core Development Team of the PMI Standard on AI in Project, Program, and Portfolio Management and served on the PMI Review Team for the PMBOK® Guide, 7th Edition. Markus holds PMP®, PgMP®, and PMI-CPMAI™ certifications and is a trainer and content creator for IIL. He delivers the course, Generative AI for Project Management, on IIL’s learning platform.
PMP®, PgMP®, and PMBOK® are registered marks of the Project Management Institute, Inc. PMI-CPMAI™ is a trademark of the Project Management Institute, Inc.
