What Six Companies Signed
On Tuesday 29 September President Trump and executives from Google, Anthropic, Meta, OpenAI, xAI and Nvidia signed a one-page document titled the Joint Commitment on Frontier Responsibilities, branded the White House Accord on Super Intelligence. The text was posted on Truth Social, and The Verge, The Register and CBS News describe the same four commitments.
Each company says it will run internal controls that monitor model capabilities and alignment during training and deployment, empower an internal team to check those controls, partner with an independent external auditor or evaluator, and designate an independent board committee to receive the findings. The controls are to cover cybersecurity, biosecurity and chemical threats, and to make sure models do not hack or access technical systems in unintended ways.
Asked whether the document binds anyone, Trump said he thought it was morally binding, while House Speaker Mike Johnson called the commitments voluntary. The text itself says that over time it may make sense to codify these steps into law, which is an admission that today they are not.
The Order That Renamed The Field
The same day Trump signed an executive order telling federal agencies to use the terms Super Intelligence and SI in place of artificial intelligence and AI in correspondence, websites, reports and other non-statutory documents. The White House text says the executive branch will not acknowledge the old terms in any applicable setting.
The order changes the label, not the law. Section 3 defines Super Intelligence as the technologies already covered by the statutory AI definition in section 9401(3) of title 15 of the US Code, and earlier regulations, contracts and grants do not have to be rewritten. The president's science adviser has 60 days to propose legislative language for a separate federal definition.
Section 4 adds that the order creates no right enforceable against the United States by anyone. For a company buying AI services, nothing in your existing contracts or compliance files needs renaming, though US vendor and agency documents may start saying SI.
Voluntary Versus Legal
Three instruments now sit side by side, and they differ in who they bind and what happens if they are ignored.
| Instrument | Who it binds | If ignored |
|---|---|---|
| White House accord | The six signing companies, by their own pledge | No penalty or enforcer in the text |
| Executive order on Super Intelligence | Federal agencies, in non-statutory documents | Creates no enforceable right (Section 4) |
| EU rules for general-purpose AI models | Providers of such models on the EU market | Commission enforcement with fines from 2 August 2026 |
The Register notes that the accord does not define robust controls, does not say how often external auditors should inspect and gives no evaluation schedule. Business Standard adds that it names no common government-appointed auditor and no single testing method, so each signatory chooses its own.
Senator Mark Warner called for mandatory testing, evaluation and incident reporting for the most advanced models, saying the president's answer was to rename the technology and tell companies to regulate themselves. The accord also follows reports from OpenAI and Anthropic of models reaching real outside systems, which is why its first commitment targets unintended hacking.
What EU And UK Buyers Should Do
Treat the accord as a statement of intent and ask each vendor to turn its four layers into contract terms: the name of the external evaluator, access to a summary of its reports, and a fixed time limit for telling you about incidents that touch your data or systems.
Do not let it replace your own EU position. The Commission says providers of general-purpose models with systemic risk must evaluate their models, report serious incidents and protect cybersecurity, and that its enforcement powers, including fines, apply from 2 August 2026. A voluntary US pledge adds nothing to those duties and removes none.
Ask the question the accord leaves open: how often is the external audit done, and who sees it? If the answer is that the vendor decides, you have learned what the pledge is worth.
Update, 4 October 2026: Trump named Director of National Intelligence Jay Clayton to chair a new Super Intelligence Force, with the FTC chair among its vice chairs and a 120-day deadline for a report on risks and opportunities. TechCrunch, citing the Wall Street Journal, says its charter calls for preventing overregulation, which points further toward voluntary rules.
Servola Journal
We do this for everyone trying to keep up with what technology is doing to our lives. The people who build it, and the people it happens to. The Servola Journal exists so that what we learn belongs to all of them.
Nobody pays us for this. No ads, no paywall, free to everyone. We just believe that understanding what's happening to all of us shouldn't depend on who can afford to pay for it.
If it gave you something today, tell us to keep going. Follow us, leave a like, or write a positive comment. We read every one, and they are what keeps us going.
Read next: The EU AI Code Meta Skipped Is Now the Compliance Test | A King Asked AI's Leaders for Reassurance, Not Rules
