AI Is Not Failing. The Operating Model Is.
Enterprise AI is not primarily failing because the models are weak; it is stalling because AI is being introduced into operating models with unclear ownership, weak governance, low-trust information, and fragmented platforms.
Executive Summary
The brief in four points.
- 01
AI value depends on the operating model that surrounds it: ownership, decisions, information, platforms, governance, and workflow design.
- 02
Gartner’s GenAI abandonment forecast and BCG’s value-gap reporting point to the same structural issue: adoption is easier than operationalization.
- 03
Large People Model interprets the gap as AI Amplification: AI accelerates the system underneath it, whether that system is clear or fragmented.
- 04
Leaders should diagnose the operating model before scaling AI, especially where accountability, information trust, platform structure, and governance are weak.
Why This Matters Now
The market shift is operational, not just technical.
AI is moving from experimentation into everyday execution. That shift raises the standard for operational clarity. Pilots can survive ambiguity because they are bounded, sponsored, and often insulated from the real workflow. Scaled AI cannot.
When an AI capability enters production, it touches ownership, decision rights, data quality, workflow handoffs, platform permissions, risk review, and adoption behavior. If those conditions are unresolved, the organization experiences the failure as an AI problem even when the deeper issue is operating model design.
This is why the market is seeing a gap between access to AI and realized value. The tools are available. The coordination system required to absorb them is often not.
“AI readiness starts with operating model readiness.”
Market Signals
Signals executives should not ignore.
30%
GenAI projects forecast to be abandoned after proof of concept
Gartner tied abandonment risk to poor data quality, inadequate risk controls, escalating costs, and unclear business value.
Source: Gartner5%
Companies reportedly achieving measurable AI value in BCG value-gap coverage
BCG value-gap reporting reinforces that AI maturity is concentrated among a small group of companies creating value at scale.
Source: BCG reporting60%
Companies reportedly seeing little to no AI benefit in BCG value-gap coverage
The value gap points toward operating-model constraints rather than simple lack of AI access.
Source: BCG reportingSignal Pattern
Gartner projected that at least 30% of GenAI projects would be abandoned after proof of concept by the end of 2025, citing poor data quality, inadequate risk controls, escalating costs, or unclear business value.
BCG reporting on AI value creation described a widening value gap: only a small portion of companies are seeing meaningful returns, while many companies report little to no benefit despite investment.
The signal is not that AI has no value. The signal is that enterprise value depends on whether AI is attached to clear work, accountable owners, trusted information, and governed execution.
LPM Interpretation
How the seven-layer model explains the pattern.
Large People Model reads this pattern through the seventh layer: AI Amplification. AI is not separate from the operating model. It reflects and accelerates the layers underneath it.
If Identity & Incentives are unclear, AI use cases lack accountable owners. If Decision Architecture is weak, teams cannot define what AI may influence or decide. If Information Ecology is low trust, AI produces output from incomplete context. If Platform Structure is fragmented, AI remains isolated from the workflow. If Governance Architecture is immature, risk scales faster than control.
The root question is therefore not “Which AI tool should we buy?” It is “Can our operating model carry the speed and autonomy this tool introduces?”
This layer is part of the operating model pattern the resource is diagnosing.
This layer is part of the operating model pattern the resource is diagnosing.
This layer is part of the operating model pattern the resource is diagnosing.
This layer is part of the operating model pattern the resource is diagnosing.
Leader Questions
What leaders should ask before scaling.
Who owns the business outcome, model behavior, workflow change, and risk?
Which decisions can AI influence, and which require human approval?
What information sources can AI trust, and who owns their quality?
Which systems must AI interact with, and where could platform fragmentation break context?
What governance controls need to be embedded before scale?
Visual Callout
AI adoption vs. operating model readiness
The LPM read: AI value is downstream of operating model readiness. When the lower layers are weak, adoption rises faster than value.
Sources / References
Reference signals used for this brief.
Gartner GenAI project abandonment forecast
Reference theme: GenAI projects may be abandoned after proof of concept because of data quality, risk controls, cost, and unclear value.
BCG AI value gap reporting
Reference theme: only a small portion of companies are achieving AI value at scale, while many realize little value.
Soft Lapemo Connection
From framework to operating system.
Large People Model defines the framework. Lapemo is being built to help organizations operationalize it across ownership, decisions, information, governance, and governed agents.