Organisational Design in the Age of AI: India Survey 2026
Indian organisations are adopting AI quickly. Rebuilding the structures around it is proving harder.
Cutting through the noise around AI is difficult. A great deal is written about its transformational potential — the threat, the opportunity, the disruption to come. Far less is known about what is actually happening inside organisations right now.
Effilor set out to answer that question. We asked a cross-section of leaders and managers not what AI might do, but what it is already doing to the way their organisations are structured, led, staffed and measured.
Over 180 responded, from technology and financial services to manufacturing, healthcare and professional services — across MNCs, Indian-headquartered firms, global capability centres and startups.
What we found was a gap. Two different speeds.
On adoption, Indian organisations are moving well. Around two-thirds of respondents report employees eager to use AI, and a similar share say AI agents already perform work once done by staff. Roles are being created as fast as they are being consolidated away.
What is moving more slowly is everything that surrounds that change.
Only a third of respondents can say who is accountable when an AI-influenced decision goes wrong. Only a third have told their people how performance will now be judged. Work is being restructured considerably faster than the apparatus around it — accountability, measurement, communication — is being rebuilt.
That distance between adoption and absorption is what this report is about.
The gap that should concern boards first
Comparing each respondent's two answers directly, comfort exceeds accountability by the largest margin of any pair in the study. Roughly half of those whose leaders are comfortable using AI in significant decisions report that nobody is clearly answerable for the outcome.
This is not primarily negligence. Organisations are building governance while discovering, in practice, that they do not yet trust AI outputs enough to remove human sign-off from anything consequential.
Three findings worth calling out
Anxiety about AI peaks not at the frontline but in middle management — the layer being asked to lead the change is also the layer whose coordinating work AI most readily absorbs.
The capabilities leaders say they most need, judgment and critical thinking, are precisely the ones whose development pathways are being de-prioritised. Coaching ranks twelfth of thirteen; resilience last.
And respondents facing similar change sort into four distinct patterns of absorption.
None of this is a technology problem. It is a leadership and organisation design problem, and it is solvable.