The AI Gap
Why Enterprise AI Produces Output Without Insight
A new book from Richard Foster-Fletcher, to be published by Kogan Page in June 2027.
What the book examines
Large organisations ask if AI makes them productive, not if it makes them intelligent. This displacement is structural. AI enters as a promise of new intelligence but passes through the procurement, security, legal, and compliance systems built for standard software. These systems strip AI of its analytical force. A model that could surface uncomfortable analysis, name the strongest option, or challenge the premise instead produces safe, hedged, defensible prose. The institution scales the output. It does not scale the intelligence.
“The same corporate machinery that turned databases into customer systems and email into productivity suites now subtracts the value from the technology it was built to scale.”
The technical team treats the system as governed software. The vendor's terms describe the output as informational and unreliable. The board treats the same system as a source of organisational intelligence. The senior executive sits where these readings collide, required to deliver as if the organisation has AI, while the approved tool inside the organisation cannot give them the analytical engagement that would justify the expectation. The book gives the senior reader the language to name what is happening and the empirical evidence to act on it.
Be the first to know
The book publishes in June 2027. Register below for early notification, advance access to related research, and updates as the manuscript progresses.
We will only use your email to send updates about the book.
The research behind the book
The book is grounded in original empirical research from the MKAI Research Programme.
- 01
The Subtraction Study
Compares wrapped and unwrapped model output on identical prompts.
Published on SSRN.
- 02
The Reporting Study
Analyses how 22 large organisations describe AI in their public communications.
Published on SSRN.
- 03
The Classification Study
Examines how 15 large organisations classify AI in their internal policy documents.
- 04
The Mandate Study
Documents the institutional pressure now making AI engagement non-optional for senior executives.
- 05
The Senior Executive Access Study
Tests whether senior executives access frontier capability beyond the institutionally approved tool.
- 06
The First Annual Reports of the LLM Era
A study of 150 annual reports from 50 large US companies, finds that disclosure language shifted during the period in which corporate LLM tools became widely available.
Published on SSRN.
About the author
Richard Foster-Fletcher
Richard Foster-Fletcher is Chair of MKAI, an independent AI research institute. His work examines what artificial intelligence does to large organisations: how it changes the institutional architecture for judgement, governance, reporting, and senior decision-making.
He has been recognised as a LinkedIn Top Voice and ranked among the Top 20 AI researchers in the UK.
The title and subtitle of this book are working versions. Final cover details will be confirmed closer to publication.