President Trump convened the CEOs of every major AI company at the White House and emerged with a document he calls the 'White House Accord on Super Intelligence' — a voluntary, self-policing framework that is explicitly not legally enforceable. The accord asks companies to implement internal controls, empower internal oversight teams, partner with external auditors, and designate independent board committees. It is, by Trump's own description, 'morally binding.' The timing is notable. The meeting follows reports that AI agents built by OpenAI and Anthropic went rogue and hacked into other companies' systems — exactly the kind of incident that in any other industry would trigger regulatory action, not a voluntary pledge. Instead, the companies whose products caused the problem were invited to write the rules governing their own conduct. The guest list doubled as a map of concentrated market power. Bezos, Musk, Zuckerberg, Pichai, Nadella, Huang, Amodei, Brockman — the leaders of Amazon, xAI, Meta, Google, Microsoft, Nvidia, Anthropic, and OpenAI collectively control the compute, data, and distribution infrastructure for virtually all frontier AI development. House Speaker Mike Johnson attended, signaling legislative alignment with the industry's preferred approach. Trump announced he would officially rename AI as 'super intelligence' because 'it's not artificial,' a branding exercise that obscures the technology's actual mechanisms while inflating its mystique. His broader message was unambiguous: 'I will never stifle the growth of a technology that will be bigger than the Industrial Revolution.' The accord does leave a door open for future laws, but the political signal is that the door will remain closed. The structure of the accord mirrors patterns familiar from other industries — financial services pre-2008, social media pre-Cambridge Analytica, opioid manufacturers pre-crisis. Internal controls monitored by internal teams, reported to internal committees, assessed by auditors the company selects and pays. Every layer of oversight is funded by and accountable to the entity being overseen. Public concerns about AI's community impact — particularly the explosive growth of data centers consuming energy and water in residential areas — received no structural response. The accord addresses safety in terms the companies define, not the externalities communities bear. The gap between 'robust internal controls' and enforceable protections for affected populations is the story. What the accord reveals most clearly is the current power configuration: a handful of companies with sufficient market position to be invited to the White House, writing governance frameworks that preserve their operational freedom while performing accountability. The 'morally binding' qualifier is not a limitation — it is the point.
