Big Tech always gets its way with this White House

IS THE White House’s artificial intelligence policy America First or simply Silicon Valley First? After this week, I’m leaning sharply towards the latter. It’s surely the only way to reconcile Washington’s conflicting policy choices, where China is made a useful bogeyman until Big Tech is interested in selling Beijing its wares. China was prominent this week in White House AI czar David Sacks’ defense of the administration’s “One Rulebook for AI,” which aims to run roughshod over states’ rights. The move is necessary to not “stymie innovation” and let China “race ahead,” he argued. Never mind that it’s also an effort to circumvent the democratic process via executive fiat, overriding the will of Congress and state legislators. But the threat of Chinese competition didn’t prevent the industry from lobbying successfully to remove a clause in a defense spending bill that would have required Nvidia to prioritize the hardware needs of its American customers over foreign sales. This opened the way, on Tuesday, for the US to lift restrictions on the export of Nvidia Corp.’s higher-end AI chips to China.

Whether or not Beijing will permit the use of these chips is an open question: Given American flip-flopping, any Chinese company building AI on the back of Nvidia hardware might wonder how long their availability will last. Still, it is a departure from earlier policymaking that hinged on the idea that American innovation should not be able to find its way to potential Chinese military use. Terrific news for Nvidia, though. And what’s good for Nvidia is good for the wider AI ecosystem, so tightly are its fortunes bound to the perception of AI’s potential. Silicon Valley First.

There was a time when decisions on major industrial policy would be a matter for elected lawmakers, debated in committees and put to the floor. Not anymore. The latest deal to allow Nvidia to sell to China was lobbied for over a Mar-A-Lago tablecloth. The One Rulebook on AI seems to have just one author, Sacks. “This is not an ‘AI amnesty’ or ‘AI moratorium,’” said Sacks, whose venture capital work has taken a backseat to his career in media and politics, in a post on X. “It is an attempt to settle a question of jurisdiction.” His attempt isn’t necessary, of course. The Senate already decided the matter when its members voted 99-1 to remove a measure designed to block state AI laws from Trump’s tax bill.

Sacks was disappointed in that result, however, and has attacked industry figures who he blamed for allowing it to happen. Deciding how AI is developed, Sacks argued, is “exactly the type of economic activity that the Framers of the Constitution intended to reserve for the federal government to regulate.” Trump said he would sign an executive order on the issue this week. As the Bloomberg Editorial Board recently stated, the patchwork of state AI laws is far from ideal. It makes building the technology more complicated and time consuming. But state lawmakers, with memory spans long enough to remember paralysis at the federal level over data privacy and social media harms, have little choice but to take matters into their own hands. Their laws have included efforts to protect minors from the harms of chatbot use amid a flurry of deeply troubling instances of AI-encouraged suicides and other harm. Others look at the risk of discrimination in the hiring process when companies use AI to pick candidates.

Sacks has been blasé in dismissing these concerns by unconvincingly arguing that existing laws can handle them. He has further poisoned the debate by making it a culture war talking point on anti-conservative bias. “The biggest threat of censorship is coming from certain Blue States,” he wrote. “Red States can’t stop this — only President Trump’s leadership at the federal level can.”

Sacks’ stated views are instructive. Saying policymaking should be centralized under a “federal framework” — note he doesn’t say “law” — is far from a commitment to enacting the guardrails the states have sought. Pre-venting the creation of these guardrails, moving fast and breaking things, is at the heart of the Silicon Valley First mindset. Leaving citizens gravely exposed to the risks is not America First. When the midterms roll around, and the presidential election after that, voters might wonder why so little attention has been paid to the impact AI might have on jobs and the economy at large.

The heavy influence of tech interests within the White House is no surprise, of course. Trump placed tech CEOs in front of his own cabinet during his inauguration, as the world looked on and wondered what the industry stood to get in return. Their tireless work to make Trump look and feel good has already offered a healthy return on investment. —BLOOMBERG OPINION