White Paper | August 2026

Governing AI in Projects:
Why the PMO Should Take the Lead

AI is already being used across project delivery. Yet in many organisations, project governance has not caught up. This paper examines why the PMO is well placed to close that governance gap, and what that responsibility means in practice.

By Emanuela Giangregorio
Download the White Paper
Governing AI in Projects: Why the PMO Should Take the Lead - August 2026 White Paper
The governance gap

AI adoption has moved faster than project governance.

The first Pulse of AI Governance in Projects report found a median maturity score of 2.43 on a five-level scale. 77% of respondents had not yet reached a standardised approach to governing AI in projects.

2.43
Median AI governance maturity score across 146 self-assessments
77%
Of respondents had not reached the Standardised maturity level
The PMO's role

Three responsibilities for governing AI in projects

The paper proposes that PMOs extend their existing project governance capability through three connected responsibilities.

01

Embed

Build AI governance into the existing project governance framework so that every AI-assisted project inherits the required controls.

02

Assure

Provide assurance that AI-assisted project work is being conducted within agreed boundaries and that appropriate human oversight remains in place.

03

Improve

Use assurance findings, lessons learned and maturity assessment to improve AI governance across the project portfolio over time.

The AIPGF connection

How the AIPGF supports each responsibility

The AIPGF is designed to extend an existing project governance framework rather than operate as a separate governance system.

Embed
Foundation, Activation and Evaluation deliverables provide practical artefacts that can be adopted and tailored as standard project deliverables.
Assure
The AIPGF lifecycle questions provide a consistent set of questions that can be incorporated into existing reviews and reporting.
Improve
The AIPG-CMM provides a structured way to assess maturity, identify priorities and track improvement across the four governance pillars.
Core artefacts
AI Assistance Plan, Data Readiness Assessment, AI Assistance Risk Register, AI Usage Report and AI Lessons Learned.
Practical starting point

Where should a PMO start?

The paper proposes practical first steps for a PMO that has not yet standardised AI governance in projects.

1. Inventory Identify AI tools and data sources already in use across projects.
2. Baseline Run the AIPG-CMM self-assessment to establish current maturity.
3. Pilot Introduce lightweight AI governance on one or two projects.
4. Assure Add the AIPGF assurance questions to existing reviews.
5. Learn Start an AI lessons repository and use the findings to improve the standard.
6. Standardise Embed AI governance into your standard project management framework so that governed AI use becomes part of normal project delivery.

Can your PMO provide assurance that AI is being used appropriately in your projects?

Read the full white paper to explore the governance gap, the PMO's proposed responsibilities and a practical route towards more consistent AI governance.

Download the White Paper

How Mature is your Governance of AI in Projects?

Ai Project Governance Capability Maturity Pillars
AI Projet Governance CMM Levels
Complete the AIPG-CMM Self-Assessment
Get certified with the APMG on the AIPGF

with APMG International

Frequently Asked Questions

Why should the PMO be responsible for AI governance in projects?

The PMO is well placed to govern AI in projects because it already has responsibility for many of the mechanisms that project-level AI governance requires: the project governance framework, standard methods and templates, assurance, and lessons learned. The paper argues that the PMO should extend this existing governance capability to cover AI rather than create a separate governance mechanism.

AI governance at organisational level may sit with an executive AI function, data governance, risk, compliance or another appropriate function. The paper does not argue that the PMO should own organisational AI policy, approved tools, ethical positions or regulatory strategy. Its focus is the application of those organisational requirements within projects, where the PMO can provide consistency, assurance and portfolio-level oversight.

The paper proposes three responsibilities: Embed, Assure and Improve.

 

Embed AI governance into the existing project governance framework.

Assure AI-assisted project work using consistent governance questions and evidence.

Improve AI governance maturity by using assurance findings, lessons learned and maturity assessment to improve the approach over time.

Embedding AI governance means making appropriate AI controls part of the organisation’s standard project governance rather than leaving individual projects to determine their own approach. This can include incorporating an AI Assistance Plan, Data Readiness Assessment, AI-related risk assessment, AI Usage Report and AI Lessons Learned into existing project artefacts, reporting and governance processes.

A PMO can incorporate a consistent set of AI governance questions into existing project reviews, stage assessments and reporting. The AIPGF provides questions across three lifecycle stages: Foundation, Activation and Evaluation. These examine matters such as approved AI use, data readiness, risks, human oversight, effectiveness, compliance with agreed boundaries and lessons learned.

The AI Project Governance Capability Maturity Model (AIPG-CMM) provides a way to assess AI governance maturity across four pillars: AI Strategy and Governance; AI Tools and Infrastructure; Human Capability and Accountability; and Data Readiness and Quality.

 

The model uses five maturity levels: Ad hoc, Initialised, Standardised, Enterprised and Optimised. A PMO can use the assessment to establish a baseline, identify priorities and track improvement.

The paper proposes a practical starting sequence:

 

  1. Inventory the AI tools and data sources already being used across projects.
  2. Establish a baseline using the AIPG-CMM self-assessment.
  3. Pilot lightweight AI governance on one or two projects.
  4. Add AI governance assurance questions to existing reviews.
  5. Establish an AI lessons repository.
  6. Standardise the approach by embedding AI governance into the organisation’s standard project management framework.

 

The purpose is to learn from practical application before establishing a more consistent governance approach across projects.

No. The paper’s central argument is that the PMO should extend the existing project governance framework, rather than create a separate AI governance system alongside it. The AIPGF is designed to integrate with existing project management approaches, including Agile, PRINCE2, PMBOK guidance and tailored in-house frameworks.

The AI Project Governance Framework (AIPGF) offers a sensible methodology for facilitating ethical, efficient and effective human-AI project collaboration.  

  • Can be integrated with a chosen project management methodology or approach, such as Agile, PRINCE2, PMBOK or hybrid approaches.
  • Provides structured and scalable AI governance, supporting projects and programmes of varying size, complexity, risk and AI adoption maturity.
  • Facilitates and encourages a high standard of ethical, efficient and effective use of AI in projects and programmes.

By implementing the Framework, organisations can systematically govern AI use across their portfolio of projects and programmes, as their AI adoption scales and as AI tools evolve.  The accompanying  AI Project Governance Capability Maturity Model (AIPG-CMM) can be used to establish maturity benchmarks and actions towards continuous improvement.

 

Disclaimer

The AIPGF is intended to provide practical guidance for governing the use of AI in projects and programmes. The author (Emanuela Giangregorio) expressly disclaims all liability to any person or organisation arising directly or indirectly from the use of, or for any errors or omissions in, the AIPGF guidance. The adoption and application of the guidance is at organisation discretion and is their sole responsibility.   

Navigation

© Copyright since 2025 Aikaizen Limited. All Rights Reserved.

 Aikaizen Limited is a company registered in England and Wales, and trades as Project Management in Practice (PMIP).

WordPress Cookie Plugin by Real Cookie Banner