Silicon Valley's political spending threatens democratic debate on AI regulation

As the huge societal ramifications of artificial intelligence have become clear, some of the tech industry’s most senior figures have taken to writing thinkpieces on what comes next. The latest to do so is Mark Zuckerberg, who this week published 6,500 words making the case for a laissez-faire approach to AI development. Worrying too much about regulation, Meta’s CEO unsurprisingly opines, could lead to the United States falling behind in the race to own the future.

Mr Zuckerberg’s plea for maximum autonomy is, of course, self-serving. Much of his essay – which airily envisions “invention superpowers” for every entrepreneur and “an abundance of jobs in the future” – reads like unadulterated tech-utopianism. But the lobbying at least has the merit of being out there in the open. As midterm elections loom in the US, darker arts are also being used to counter popular concern about what AI will mean for a host of topics, from the labour market to child safety, the environment and cybersecurity.

Replicating a strategy deployed by cryptocurrency backers in the 2024 election cycle, Silicon Valley is pouring hundreds of millions of dollars into Super Pacs – groups unaffiliated to political parties or campaigns – to ensure that voices sympathetic to its deregulatory AI agenda get elected. OpenAI’s co-founder and president, Greg Brockman, and Palantir’s co-founder, Jo Lonsdale, have given large sums to Leading the Future, an organisation dedicated to “identifying, maintaining, and growing pro-AI candidates”. Vast amounts are being spent to try to tilt the balance in individual contests, sometimes through disingenuous means.

Earlier this summer, Leading the Future spent $6m on ads attacking Alex Bores, a New York state assemblyman and pioneer of AI regulation at state level. Rather than focus on his AI campaigning, the ads attacked Mr Bores’s progressive credentials, alleging that he had carried out tech work in the past for US Immigration and Customs Enforcement, which he denies. Running for Congress in a Democratic stronghold, he lost a tight primary race after battling what was in essence a smear campaign.

Countervailing forces are emerging. Guardrails Alliance, a union-backed Super Pac, is attempting to combat Leading the Future’s influence, while a bipartisan groundswell of opposition to datacentres is generating serious electoral clout. But as tech companies and their billionaire owners seek to shape and above all limit the scope of AI regulation, their influence is distorting debate in a political culture where the power of money already holds far too much sway.

In his essay, Mr Zuckerberg urges that AI’s democratising potential be protected from those who believe “a small set of experts … should decide what is best for humanity”. He appears to miss the irony that an unelected tech oligarchy, to which he belongs, is already performing that function by default. Its pressure has been instrumental in persuading Donald Trump to cancel the signing of an executive order which would have given the government the right to review new AI models. Its wealth is now being used to attack and undermine dissenting voices in the public square.

If a democratic debate about the benefits and dangers of AI is to take place, such presumption must be fiercely resisted in the US and elsewhere. The semi-covert manipulation of political discussion is no way to map out an inclusive future in challenging times.