NORBERT BIEDRZYCKI

OpenAI and the consultants: the last mile of AI is organizational

When OpenAI announced partnerships with McKinsey, BCG, Accenture and Capgemini, it didn’t feel like a breakthrough moment.

At first glance, it looked like another predictable move: big tech teaming up with big consulting. But the more you think about it, the more it reveals something deeper about where we really are with AI.

For the past two years, the conversation has been dominated by capability. Better models, better outputs, more powerful tools. And to a large extent, that problem is already being solved. The technology works, and it is improving faster than most organizations can realistically absorb.

And yet, inside most companies, very little has actually changed. AI pilots are everywhere, but real, scaled implementations are still rare. Teams experiment, run proofs of concept and generate excitement, but struggle to move beyond that. The issue is no longer access to AI. It’s the gap between what the technology can do and what organizations are able to implement.

Companies are not built to absorb something like AI overnight. Their processes are rigid, their data is fragmented, and their decision-making assumes a level of certainty that AI simply doesn’t provide. You can’t just “plug AI in” and expect it to work. It forces changes in how work is structured, how decisions are made, and who is accountable for the outcome.

Instead of trying to solve this internally, OpenAI is leaning on players who already operate inside large organizations. Consulting firms, for all their imperfections, have one major advantage: they understand how companies actually function. They have relationships, access, and the ability – at least in theory – to drive change from within.

In that sense, they are becoming something more than advisors. They are becoming a distribution layer for AI. Not for ideas, but for execution.

This shift matters, because it changes where value is created. For a long time, attention has been focused on the model layer. But as models become more accessible and more standardized, differentiation starts to move elsewhere – into the messy space between capability and real-world application. Integration. Workflow design. Decision ownership. Governance.

For private equity, this creates a very practical opportunity. AI can drive real operational improvements across portfolio companies, but only if it’s implemented in a structured, repeatable way. Buying tools is easy. Changing how companies operate is not.

For venture capital, the signal is slightly different but just as important. The next wave of value will not come only from building better models, but from making those models usable inside real organizations. The “last mile” of AI is still wide open.

The uncomfortable truth is that AI doesn’t simplify organizations. It makes them way more complex.

Source: Fortune

Read next

  1. Solow's déjà vu: AI everywhere except in the productivity statisticsLinkedIn

    I've been participating in technology revolutions since the early days of my career, and in AI since the moment it became deployable at scale in real commercial environments. And yet, I increasingly feel a sense of déjà vu.

  2. AI was supposed to get rid of consultants. It's changing what they doLinkedIn

    At first glance, it made sense. If a model can analyze data, generate insights and even suggest a strategy, why would companies still need external advisors? But that assumption is starting to crack. In practice, most companies don't struggle with access to AI.

  3. AI won’t run out of customers. It will run out of power and time to repay its debt

    Many people read the sell-off in data center stocks as the first crack in the AI bubble. I look at the same numbers and see something else. The demand is there, and it is substantial. The trouble lies elsewhere: in the construction itself, in access