Moderna: the right instinct, the wrong structure

In 2025 Moderna did something no large company had done. It merged human resources and information technology into a single function, People and Digital Technology, and put its former chief human resources officer, Tracey Franklin, in charge of both. The company had built more than 3,000 custom GPTs with OpenAI and expected its systems by 2030 to include humans and AI agents working side by side. Planning people and technology separately, Franklin argued, no longer made sense.

I think the instinct is right and the structure is wrong, and the difference matters for every company that will be tempted to copy it.

The instinct is right because the separation of workforce planning from technology planning has become untenable. When a technology decision is a workforce decision, as deploying an agent now is, having two functions plan independently guarantees that one of them discovers the other's decisions after the fact. Moderna saw that clearly, earlier than most. Franklin's phrase for the goal, "work planning" rather than workforce planning and technology planning, is exactly the right phrase.

The structure is wrong because a merger buys coordination at the price of judgment. Consider what the merged function now contains. On one side, the people who select, build and deploy the systems, whose budgets carry the systems' costs and whose success is measured by their adoption. On the other, the people who protect, develop and represent the workforce, whose concerns show up later and less legibly: a thinning of expertise, a loss of trust, the disappearance of the entry-level roles through which people used to learn. Put both under one leader with one budget and every tension between them is resolved inside one head, under the pressure of a quarterly plan. The resolution will not always favour the system. But the structural pull is in that direction, because the system's benefits are visible this quarter and the human costs arrive in year five.

There is a reason organisations do not merge the chief financial officer with the head of sales, even though they must coordinate constantly. Each must be able to say no to the other, in front of the chief executive, and the organisation needs to hear the disagreement. The same logic applies here. The leader who speaks for the workforce and the leader who builds the systems should be peers who argue, not one person who decides.

Franklin herself describes the merger as "still very much a work in progress," and I take that seriously; Moderna may well find structures within the merged function that preserve the independent voice. But most companies that copy the headline will not copy the care. They will merge the two functions, call it transformation, and quietly lose the only executive whose job it was to ask whether a cheaper and more capable agent should still not be deployed where nobody can check its work.

The alternative is not the old separation. It is a Chief Workforce Officer, accountable for the whole productive capacity of the organisation, human and artificial, who sits beside the chief information officer rather than above, below or merged with that office. The CWO decides what work agents do and who answers for them; the CIO decides how the systems are built and run; and when they disagree, which they will, the disagreement reaches the top table intact.

Moderna asked the right question first. The rest of us have the advantage of answering it with a structure that keeps the argument alive instead of settling it in advance.