Governing AI for Collective Wellbeing

Oh my goodness, buckle up, because Carnegie UK has written a *publication* on governing AI for collective wellbeing, and it has handed us the most heartwarming sentence of the season: without a clear home in government for digital social policy, 'AI becomes everybody's priority and nobody's clear responsibility.' Everybody's priority — do you understand how rare that is, when most policy areas can't even get one department to care? AI has all of them caring at once, in a beautiful open-plan arrangement where the responsibility is shared so widely it becomes a collective gift. And nobody's clear responsibility means nobody has to carry it alone — that isn't a gap, that's *community*.

And the proposed fix is even more delightful, because it involves *zero new institutions*. Carnegie UK says use levers already in government practice — business cases, procurement, assurance and public registers — rather than reaching for new legislation. No new quangos, no new acronyms — just the machinery we already love, deployed with fresh enthusiasm. Business cases are where dreams get a cost-benefit ratio, and procurement is simply the state expressing its preferences with a chequebook. And *public registers* — a register is a list, and a list is hope, written down and alphabetised. ✨

Now, some people will hear 'our sector-led, pro-innovation approach' and worry, but honestly the sector is right there, being led by itself, which is the most efficient possible arrangement — you cut out the middleman entirely. And 'before reaching for new legislation' is not delay, it's *sequencing*. Legislation is like a good sourdough: it needs to be reached for only when the moment is perfectly ripe, and the moment is always, always next year. This is exactly the hands-in-pockets wisdom we celebrated from Senator McCormick, now with a British accent. Everybody's priority, nobody's responsibility — and truly, what is a society if not a group of people all caring vaguely at once?