Sanders and Bannon push AI curbs at Pro-Human Assembly

At Washington’s Pro-Human Assembly, Sen. Bernie Sanders and Steve Bannon both pressed for tougher AI guardrails. They agreed Silicon Valley needs restraints, then split on China — diplomacy and a superintelligence ban versus economic and tech quarantine.

Same-day POLITICS desk: unusual left–right convergence on AI restraint at a Washington event, with concrete competing policy asks — a pause and superintelligence ban versus a China tech quarantine — as the industry safety debate intensifies.

On Tuesday 15 Sep 2026, the Pro-Human Assembly convened in Washington, D.C. Sen. Bernie Sanders, the independent from Vermont, and Steve Bannon, the former Trump White House strategist, both spoke. Bloomberg’s 15 Sep 2026 dispatch, plus same-day CBS, Guardian, and CNN coverage, is the filing event. These are on-record remarks at a public event. They are not a new law, not a House vote, and not a White House agreement. The Guardian and CNN both say Bannon spoke after Sanders but did not share a stage with him. This desk did not sit in the room.

They appeared with a bipartisan group of members of Congress pressing a “human-centric” approach and skepticism of the AI companies leading the industry, CBS reported. Human-centric here means putting people first when building and releasing AI. File that framing as CBS’s. This desk is not treating it as enacted policy.

The Guardian reported that Sanders and Bannon were united in warning about AI dangers and demanding stringent guardrails against Silicon Valley “oligarchs,” but offered competing visions on what they termed a “cold war” with China. Guardrails here means rules and limits on how AI is built and released. File that unity-then-split as the Guardian’s. This desk is not inventing a joint Sanders–Bannon bill.

Sanders, as the Guardian reports his speech, called for international diplomacy and a global agreement, including with Beijing, to ban development of super-intelligent machines. Superintelligence here means AI smarter than humans across most tasks. He noted President Trump is due to host Chinese President Xi Jinping in Washington next week, and said he hoped those talks would negotiate a pause on advanced AI development and a ban on superintelligence. File those asks as Sanders’, via the Guardian. This desk is not saying a treaty was signed or that the White House adopted that agenda.

CBS quotes Sanders opening: if AI surpasses human intelligence, as many scientists believe could happen, the technology will escape human control with potentially catastrophic consequence. He also said AI is advancing exponentially every single day, so current systems are the least capable version of AI humanity will ever have. File those lines as Sanders’, via CBS. This desk is not independently certifying the science.

Bannon spoke after Sanders but did not share a stage with him, the Guardian reported. He used cold-war framing toward an aggressive economic and technological “quarantine” of China. Quarantine here means cutting off tech and economic ties, not a medical quarantine. File that China ask as Bannon’s, via the Guardian. This desk is not treating quarantine as enacted policy.

CBS: Bannon, taking the stage after Sanders, applauded the senator’s speech. He said they may disagree politically and on President Trump, but they are “trying to rise above partisanship, and we have to.” He blasted tech CEOs who say they want more regulation as a “bald-faced lie,” claiming they want to “socialize” the risk, shut out competitors, and make a “liability shift to the American people.” File those lines as Bannon’s, via CBS.

Context only, as CBS reports it: on Saturday, Anthropic CEO Dario Amodei called for a slowdown at frontier AI labs, quickly echoed by OpenAI CEO Sam Altman and Elon Musk. Those weekend remarks are already filed as amodei-pace-the-frontier and altman-musk-hassabis-back-amodei-pacing. This filing is Tuesday’s Washington event only.

Plain English for the rest of the card: superintelligence = AI smarter than humans across most tasks. Guardrails = rules and limits on how AI is built and released. Quarantine here = cutting off tech and economic ties, not a medical quarantine. Human-centric = putting people first when building and releasing AI. Pause = a freeze on building the next advanced systems.

PRIMARY here: Bloomberg’s 15 Sep 2026 independent Tier B dispatch — plus same-day Guardian, CBS, and CNN event coverage of the same public Washington gathering. The Pro-Human Assembly date and place, both men speaking, the human-centric / company-skepticism framing, the oligarchs / China-split, Sanders’ diplomacy and superintelligence-ban ask, the Trump–Xi hope, the CBS Sanders quotes on escape-from-control and exponential advance, Bannon speaking after Sanders without sharing a stage, the China quarantine, Bannon’s applause / bald-faced-lie / socialize-risk lines, and the weekend Amodei / Altman / Musk context are Bloomberg-, Guardian-, CBS-, or CNN-attributed as named. NOT claimed: a bill number, that a law passed, White House agreement, that Sanders and Bannon shared a stage or co-hosted, joint Sanders–Bannon legislation, a Senate.gov / official transcript primary, or that this desk sat in the room. Distinct from the already-filed mike-johnson-no-ai-moratorium, google-org-ai-gov-challenge, and china-tc260-ai-safety-framework-3-0.