Claim
Enterprise AI vendors face a structural prisoner's dilemma: correcting a customer executive's inflated AI productivity claim would undermine that executive's credibility, so vendors stay silent, locking the ecosystem into a self-reinforcing hype cycle with no internal correction mechanism.
Mechanism
When a customer executive publicly commits to 100x productivity projections, any vendor contradiction invalidates a claim the executive has staked their credibility on. The vendor risks losing the contract; the customer risks losing face. Both parties therefore sustain the framing even when both know it is implausible. The dynamic scales across the ecosystem: executives at competing companies feel pressure to match peers' claims, and vendors with relationships at all those companies cannot correct any one without endangering all others. Demo psychology compounds the problem: Suresh documents clients who were initially lukewarm becoming urgent buyers after watching a product demo, regardless of honest capability warnings given in the same session.
Conditions
Holds when:
- The customer executive has publicly committed to specific AI productivity targets in front of their own leadership.
- The vendor's revenue is tied to that customer's continued belief in the projection.
- No independent third party (regulator, auditor, board) requires verifiable productivity data before renewal.
Fails when:
- Procurement requires audit-ready evidence of AI ROI before contract renewal, forcing honest accounting.
- An executive has internal cover (budget surplus, board trust) to revise targets without career risk.
- The vendor holds a dominant market position and can afford to lose the contract without commercial consequence.
Evidence
"Executives at their customers were saying absurd things about achieving 100x productivity, and this meant that if any executive at the vendor said that these gains were not plausible, it would undermine the credibility of the customer's executive."
Suresh adds a specific observation from consulting practice: "an executive confess that they had never even used ChatGPT or any AI tool in their life, immediately after producing a technical strategy for an organisation with $2B+ in revenue which was entirely centered around AI."
Signals
- Your AI vendor cannot give you a realistic baseline for productivity gains without referencing their own marketing materials.
- Your organization's AI strategy cites projected numbers rather than piloted numbers.
- Internal AI champions quote vendor case studies instead of internal data when justifying spend.
- No one in the vendor meeting says "that target is not realistic."
Counter-evidence
Not all vendor relationships work this way. Management consultancies and auditors sometimes deliver unwelcome findings without losing clients, particularly where fiduciary duty applies. Some enterprises have built internal AI evaluation teams that require empirical ROI data before committing budget to scale.
Cross-references
- (none in current corpus)