Breaking

Trump and Six AI Leaders Sign Voluntary 'Super Intelligence' Safety Accord

September 30, 2026 at 13:29 EDT

Policy & Regulation
Foundation Models
Security

President Donald Trump and six top AI executives signed a voluntary White House Accord on Super Intelligence on September 29, 2026, setting out a four-layer governance plan for companies building frontier models. The agreement pairs industry self-policing with a separate executive order directing federal agencies to use "Super Intelligence" and "SI" in place of "Artificial Intelligence" and "AI" in most executive-branch communications.

September 29, 2026 · Federal AI policy

A voluntary safety accord for "Super Intelligence"

President Trump and six AI leaders backed a four-layer control framework while federal agencies were told to use SI terminology in most official communications.

executive signers from major AI and chip companies

layers of controls and audits in the accord

days for proposed SI definition language

days for America.gov implementation guidance

Two deadlines, drawn to scale

60 days

SI definition proposal

90 days

America.gov guidance

Each block represents 10 days.

The accord's control flow

Internal controls

Monitor capability and alignment risks.

→

Internal team

Verify monitoring and remediation.

→

External evaluator

Assess whether controls operate as intended.

→

Board committee

Receive reports and oversee fixes.

Why supporters like it

Shared safety norms across leading labs.
External evaluation enters the formal control chain.
Regular meetings could align best practices.

What remains open

The accord is voluntary, not legally binding.
No auditor or public reporting duty is named.
Congress could still pursue mandatory rules.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.