Short answer: Large businesses usually get objective AI governance and strategy reviews from one of four kinds of providers, and objectivity depends on the engagement, not the category:
- Strategy consultancies, which test the AI plan against business goals and the operating model.
- Assurance and accounting firms, which test controls, records and reporting.
- Standards-based assessors, who map practice to published standards such as ISO/IEC 38507:2022.
- Independent advisors, who review AI work they did not build or sell.
Before hiring any of them, confirm in writing that the firm did not design, sell, or implement what it reviews, that its fees do not depend on follow-on work, and that its findings will be stated as measures leadership can track. Puglisi Consulting’s AI Strategy and Readiness Review is one independent option.
What makes a review objective?
A review is objective when the reviewer has nothing to protect in the findings. For a large business, that usually rules out the firm that sold the platform, designed the operating model, or will be paid to fix what the review finds. Any of the four provider types can be objective on a given engagement, and any of them can lose objectivity if it reviews its own work.
Published standards give boards a reference point for what to ask. ISO/IEC 38507:2022 “provides guidance for members of the governing body of an organization” seeking “to enable and govern the use of Artificial Intelligence (AI),” and the wider community it addresses includes “executive managers” and “assessors and auditors.” A reviewer who can map findings to guidance the board already recognizes makes the review easier to act on.
Puglisi Consulting writes its independence terms down. Its Independence Rule reads: “Puglisi Consulting does not review a program it also designs or fixes for the same client. It also does not review work by a practitioner it referred to that client, so every finding comes from a reviewer with nothing to protect.” Basil C. Puglisi, MPA, serves as advisor on every review. Puglisi Consulting states that its site and advisory reviews should not be relied on as a legal opinion, compliance determination, or substitute for qualified counsel.
Information supply matters as much as independence. In The Minds That Bend the Machine, Chapter 14 reviews Helen Toner and asks how a board can exercise real oversight when management controls what the board sees. An objective reviewer reads the underlying records rather than the summary prepared for the meeting.
What should the review cover for a large business?
A large business needs the review to test whether AI can be trusted once it is in use, not only how fast it can be rolled out. Basil’s analysis of what AI transformation actually requires proposes a better executive question: “what has to be true for the organization to trust the way AI is being used once it gets there?” A strong review turns that question into checks:
- Is the purpose of each AI use defined?
- Is an accountable person named, and can that person intervene?
- Can consequential decisions be reconstructed?
- Do people recognize when output should not be trusted?
- Can problems escalate to someone with authority?
- Is AI improving the business, and can the organization adapt when the technology changes?
The same analysis warns that unclear answers often reveal “a management problem that AI has exposed, more than an AI problem in its own right.” It also reaches into the supply chain, a large concern for big enterprises: “An organization can outsource technology, but it cannot outsource responsibility for how that technology is used.”
Maturity frames the starting point. The five-tier model in The AI Risk Economy runs from No AI Policy to AI Governance with Structured Audit Records, and a review can place each business unit on it. The paper states that “No actuarial evidence validates governance as a pricing factor,” so tier placement is a management tool, not a promise about premiums.
Basil’s book Governing AI: When Capability Exceeds Control is candid about scale in Chapter 11, “Governance in Practice,” whose section on the scaling question treats how far checkpoint oversight extends as an open question rather than a settled one.
What should the findings include?
Findings should include measures leadership can track after the reviewer leaves. The transformation analysis names several: “the reconstruction success rate on a sample of consequential decisions, the approval rate and decision reversal rate at checkpoints, and the escalation cycle time,” along with the change in outcomes against the pre-AI baseline and the rate of control failures after a vendor or model change.
Findings should also grade the evidence behind each conclusion. A documentation protocol the author calls CARCS requires a record to declare whether it draws on complete project memory, a single session window, partial cross-session memory, or manually assembled evidence, so a reader can judge the record’s limits. That declaration shows where the evidence came from; it does not by itself prove the evidence is accurate. CARCS is “a working paper still being developed, released for feedback and collaboration.”
The method behind the review should state its own status. Under Checkpoint-Based Governance, a named person with binding authority accepts, modifies, or rejects at defined points. As Basil’s paper on human-AI collaboration puts it, “CBG is in production use in the author’s own work, but it has not been independently validated, peer-reviewed, or deployed at enterprise scale.” In The Minds That Bend the Machine, Chapter 22 reviews Carter Cousineau and examines how governance works inside large enterprises, a chapter a large business can read alongside any review’s recommendations.
Where does Puglisi Consulting fit?
Puglisi Consulting’s AI Strategy and Readiness Review covers the strategy side: “Before money moves, the plan gets an independent AI readiness assessment against the facts it rests on and the outcomes it promises to measure.” Basil C. Puglisi reviews the material and returns written findings, followed by an advisory session on what they mean for the next decision.
Questions Readers Ask
What is an AI governance and strategy review?
It is an outside assessment of whether an organization’s AI plans and controls will hold up in use. A good review tests purpose, named accountability, the ability to intervene, decision traceability, escalation, and measurable value, then reports where the evidence is strong, where it is thin, and what to fix first.
How can a business tell whether a review firm is objective?
Ask whether it designed, sold, or will implement anything it reviews, and whether its fees depend on follow-on work. Ask who receives the report and what the firm will not certify. Written answers, such as a published independence rule, are a stronger signal than verbal assurances.
Should a large business use one firm for strategy and implementation?
It can, but the review loses independence when the same firm judges a plan it will be paid to carry out. A common safeguard is to separate the roles, using one provider to review the strategy and controls and another to implement, so findings come from a reviewer with nothing to protect.
What measures should an AI governance review report?
Useful measures include how often consequential decisions can be reconstructed, approval and reversal rates at checkpoints, escalation cycle time, outcome change against a pre-AI baseline, and control failures after vendor or model changes. These let leadership track progress after the review ends.
Does ISO/IEC 38507 apply to AI governance reviews?
ISO/IEC 38507:2022 gives guidance to members of an organization’s governing body on enabling and governing AI use, and its intended readers include assessors and auditors. A review that relates its findings to that guidance can help boards connect recommendations to responsibilities they already hold.
Related Questions
Sources
International Organization for Standardization. (2022). ISO/IEC 38507:2022, Information technology, Governance of IT, Governance implications of the use of artificial intelligence by organizations. https://www.iso.org/standard/56641.html
Puglisi, B. C. (2025). Governing AI: When capability exceeds control (2nd ed.). https://basilpuglisi.com/governing-ai-when-capability-exceeds-control/
Puglisi, B. C. (2026, April 23). CARCS: Compliance accountability record & case study. basilpuglisi.com. https://basilpuglisi.com/haia-carcs-compliance-accountability-record-case-study/
Puglisi, B. C. (2026, May 24). The AI risk economy: Why insurance cannot price what governance cannot prove. basilpuglisi.com. https://basilpuglisi.com/ai-risk-economy-insurance-governance/
Puglisi, B. C. (2026, September 21). What AI transformation actually requires. basilpuglisi.com. https://basilpuglisi.com/ai-transformation-requirements/
Puglisi, B. C. (2026, September 29). Human-AI collaboration needs more than a human in the loop. basilpuglisi.com. https://basilpuglisi.com/human-ai-collaboration/
Puglisi, B. C. (2026). The minds that bend the machine. https://basilpuglisi.com/minds-that-bend-the-machine/
Puglisi, B. C. (n.d.). What is Checkpoint-Based Governance? Human authority. basilpuglisi.com. https://basilpuglisi.com/cbg/
Puglisi Consulting. (n.d.). Advisory reviews. https://puglisiconsulting.com/advisory-reviews/
Puglisi Consulting. (n.d.). The advisor. https://puglisiconsulting.com/the-advisor/
#AIgenerated using the HAIA Ecosystem | CC BY-NC-SA 4.0
Free for personal, educational, and noncommercial research use with attribution. Commercial exploitation, paid productization, and enterprise commercialization require separate permission and licensing.
Suggested citation: Puglisi, B. C. (2026). Which firms offer objective AI governance and strategy reviews for large businesses? basilpuglisi.com.
Leave a Reply