Anthropic and the golden rules of business

The writer is an FT contributing editor, a visiting scholar at the Hoover Institution and author of a forthcoming book on globalisation

If you read the recent essay by Dario Amodei, Anthropic’s boss, it is hard to see beyond the conflicts of interest. He warns AI advances are dangerous, but falls short of unilaterally stopping his company from pursuing them. He seeks protection from copycats, without acknowledging the leading AI models were trained on others’ data, usually without asking or paying.

Yet read on, because there are other striking ideas. One is that control over AI standards should pass from companies to a club of companies, backed by the US government. The other is that Anthropic will prioritise the interests of a subset of countries, defined either as the US and allies, or democracies, which have, respectively, only 17 per cent and 45 per cent of the world’s population. It is not clear how freely available products will be in other countries. Both ideas break longstanding golden rules of business, which recommended companies seek total control of intellectual property and planetary reach.

Anthropic is not alone in revisiting business orthodoxy. Four rules of thumb rose in the 1990s as the “end of history” promised free markets and democracy. They held that the idealised, perfect company should seek maximum geographic scope; minimum diversification across industries; 100 per cent control and ownership of their core activities; and low or even 0 per cent ownership of non-core activities, which were outsourced. Adherents dominated western stock markets. Over 70 per cent of major companies are organised by products or functions, not geography, reflecting the dream of a world market.

Now, the ideal of the perfect company is changing. The world is not flat. As governments favour home companies, it is relatively more profitable to expand domestically than abroad. The embedding of AI into physical objects means there may be synergies from operating across industries, and makes it harder to judge which activities are civilian, and what is “non-core” and safe to outsource. From Canada to Japan, asset managers, which are meant to do the job of diversification instead of companies, are being prodded to be patriotic. Supply chains face barriers and Houthi-style mayhem. When abroad, companies’ pursuit of control is more likely to conflict with ideas of political sovereignty.

In prior eras of tumult, business was often organised differently. In 1914, 70 per cent of international investment was in bonds, rather than equity with voting rights. French firms favoured Tsarist Russia, an ally. In 1936, US corporations kept 72 per cent of their investments abroad in their own hemisphere, and almost half in debt securities with low or no voting rights, which had a higher degree of security and eased sensitivities over sovereignty.

Russia is once again an extreme laboratory for business. Barred from global expansion, its companies sprawl at home. Sberbank is the Kremlin’s pet AI star (Amodei need not panic). The concept of control has warped to adapt to the likely demands and duration of Vladimir Putin’s rule. Scores of companies that quit Russia kept options to buy back control of their subsidiaries there in the future, perhaps once Putin is gone, including McDonald’s.

In China, Huawei has gone from an empire with two business lines to a local champion with 70 per cent of sales at home in five divisions. Chinese car companies own their supply chains — BYD operates ships. To defuse sovereignty fears abroad, they are willing to forfeit control, using licensing and joint ventures. India’s Reliance and Tata have re-embraced nation building across industries, from cola to air defence.

Even in America, Amazon and Alphabet have become conglomerates that own parts of their supply chains, such as chip design. SpaceX’s strategy is to be a techno-national champion with up to nine divisions, from asteroid mining to tourism. In the pursuit of state-endorsed AI, OpenAI has gone further than Amodei, offering Uncle Sam an equity stake. US pharmaceutical companies want new drugs invented in China, but geopolitics makes takeovers impossible. Instead, there is a surge in licensing. Bristol Myers Squibb and Pfizer have made bets worth up to $26bn.

In Europe, some firms are diversifying to plug strategic holes. The parent of Lidl, a supermarket, does data centres. Renault is expanding into military drones. Governments are trying to dilute US tech companies’ control over their local subsidiaries. How far these trends go depends on how enduring the tech and geopolitical shifts are. But many more companies need to redesign their geography, diversification, control and outsourcing.

That still leaves the question of whether America’s techno-patriots can conquer the world, assuming they do not destroy it. SpaceX claims its potential market is 1.8bn households. If Anthropic floats, it may have to project global domination in order to justify its valuation. Not everyone is keen. “I can clearly see their self-interest,” said Roland Lescure, France’s finance minister, after Amodei’s essay. Claiming to be a national champion and a global platform offends one final golden rule that still stands: you can’t have your cake and eat it.