The flow of a governance crisis often follows Ernest Hemingway’s famous dictum on how someone becomes bankrupt: gradually, then suddenly.

Personal peccadilloes, outsized voting arrangements, erratic social media outbursts – all can lay the groundwork for potentially explosive dust-ups that destroy trust, then value.

Nvidia CEO Jensen Huang is certainly not the first man of a certain age to don a leather jacket and plunge head-first into the murky waters of indiscretion. Such personal foibles rarely become an issue of professional concern, unless exposed at a Coldplay concert.

But Huang’s acknowledgement of an extramarital affair with a data centre executive suspected of being a Chinese spy is particularly consequential for a driving force of the AI revolution and leader of a USD 4.3tn chip behemoth.

It’s also a reminder of the huge “key player” risk attached to so many AI executives: Elon Musk at SpaceX and Tesla; Sam Altman at OpenAI; Mark Zuckerberg at Meta.

And, of course, Anthropic’s co-founder siblings Dario and Daniela Amodei.

Today (10 November), Anthropic filed its prospectus for an IPO that will see the siblings not only sell shares at the listing – a brave signal amid wider market ructions – but also retain key control alongside other co-founders of 50.1% of the group’s voting rights via a single Class F share.

Anthropic’s IPO valuation of up to USD 1.4tn, on a fully diluted basis, looks punchy versus analysis undertaken by this column’s equity issuance-focused twin ECM Pulse, although its non-diluted value of around USD 1tn is more in line with that number crunching.

The IPO price might seem particularly rich given the value already torched in a torrid few weeks for AI-focused, capex-heavy stocks: Nvidia is down 21% this month, Alphabet 18%; CoreWeave 39%.

Concerns among bears that AI valuations had overheated ahead of a likely price war – and with local inference potentially limiting return on investment in frontier models – were already in the mix before a Democratic sweep in US midterms curbed bulls’ enthusiasm by raising fears that regulatory roadblocks will slow data centre rollout.

From land use through to energy and water consumption, widespread citizen revolts against data centres have sprung up in the US across red and blue states alike, with Democrats thus far most effective at channelling this activism into votes.

A poll commissioned by AP and others found 56% of US voters cited data centres as a “crucial or major” factor in the midterms, with 76% adding they would oppose proposals to build data centres in their communities.

Alongside fears about job erosion and the potential eradication of the human race, the collection of risks to be managed if AI is to reach its full potential might best be summarised as: environmental, social, and governance.

Once an omnipresent, overarching prerequisite for any M&A, IPO or financing discussion, ESG has all but become a punchline since President Donald Trump’s re-election in November 2024. Flowing from the change of mood in Washington DC, a vibe shift across global investment and dealmaking dented capital’s appetite for ESG as a strategy.

In FY25, global sustainable funds experienced USD 84bn in net outflows versus USD 38bn in inflows over FY24, according to Morningstar.

Some of the backlash was justified in Continental Drift’s opinion. ESG taxonomy efforts to direct capital long felt obscurantist. “ESG fell apart because it became too thinly defined – a sort of ragbag of ‘things that I like’ from whoever had the definitional power,” one cross-border dealmaking lawyer said.

The scramble to redefine defence investment within ESG parameters after Russia’s full invasion of Ukraine in 2022 was a case in point.

But as corporate practice, ESG is absolutely critical to demonstrating the kind of transparency and sustainability – in every sense of the word – that investors rely upon to make informed capital allocation decisions.

Governance – something of a runt of the litter even in ESG’s pomp – is particularly critical in this context. After all, with its huge existing compute capacity, SpaceX ought to be well positioned if data centre build-out slows. Yet Elon Musk’s outspoken partisanship has put his business interests in Democratic crosshairs ahead of the new Congress sitting next year.

Perhaps this market moment represents an inflection point.

“The great irony of the ESG backlash is that it may accelerate the transition from ESG as an ideology to ESG as an investment discipline and economic necessity,” according to Michel Driessen, Chair at Queen’s Tower Advisory, Director of the M&A Research Centre at Bayes Business School, and co-author of Green Gold: How Sustainability Creates Deal Value.

Nowhere is this truer than in the world of AI, where public consent will increasingly condition the speed and scale of AI development and adoption.

“Energy and water are – and will become – scarce resources. This will move the investment case from a pure ESG perspective to an economic and financial perspective,” Driessen argues.

Executives may baulk at the idea of sitting round the campfire singing Kumbaya – but laying out a convincing narrative on issues traditionally considered core ESG terrain will become increasingly mainstream once again.

“The financial and operational implications are too significant to ignore,” Driessen says.

Perhaps Nvidia’s board will overlook its CEO’s dalliance. In a statement following revelations of Huang’s affair, it said it had launched an independent investigation into the matter, while asserting “the Board has no reason to believe that either national security or Nvidia secrets were compromised”.

Very reassuring.

Perhaps investors will also ignore Anthropic’s rich valuation and concentrated voting rights to gain exposure to one of the indisputable leaders in frontier AI models, regardless of its massive irrevocable compute spend commitments.

But sooner or later, a hard rain is going to fall on companies that fail to provide investors or suitors with compelling visibility on the impact of their business plans given swelling dissatisfaction with corporate actors both sides of the pond.

As Hemingway also wrote: “the way to make people trustworthy is to trust them”. In the case of the AI Prometheans, we’re about to discover if that trust was misplaced.

Continental Drift is a weekly column offering commentary on the macroeconomic, political, and policy forces shaping the M&A landscape across the US and Europe. The opinions expressed here are those of the writer only.