The Race — Building Democracy in the Age of AI. People race toward a shared finish line.
The Race
← Trends

Established trend · Active

AI implementation expertise is organized through embedded provider practices

Latest assessment: · Evidence updated

Standing score
Unscored
Direction of power
Leaning concentration
Window
Unclear
Confidence
Moderate

What we think is happening

Integrators are naming and staffing client-embedded teams that combine model expertise with domain/workflow redesign. TCS, Accenture/Google, Cognizant and Wipro show implementation becoming a structured service offering whose knowledge may remain outside the client.

Scope: Large IT-services firms; recurring organizational practice, not a measured new gatekeeper market.

Potential direction of power

Provider teams increasingly supply the expertise needed to implement AI. Much of this adapts an existing consulting model, and staffing targets alone do not establish lasting dependence. The key question is whether clients retain the knowledge, work products and ability to operate after the engagement.

Supporting cases

Window of contestability

Unclear for clients over implementation/renewal: embedded expertise can teach institutions or leave necessary knowledge with the integrator.

What would change this assessment

Client ownership of work products, effective training and documentation, manageable exit costs and successful operation after handover would show whether implementation support builds independence.

Related cases

← All 24 trends