1POP unveils global AI enforcement plan with veto power and breakup penalties

One Planet – One People released its Global Protection Resolution on Sept. 29, 2026, laying out a proposed international AI safety regime with pre-release testing, binding rulings and escalating penalties. The plan arrives as the UN debates AI governance and as 1POP pushes for stronger global rules on frontier AI and autonomous weapons.

Why it matters: - One Planet – One People is proposing a binding global enforcement model for frontier AI, not just voluntary guidelines. - The framework would let a new international commission block unsafe systems before release, which could reshape how advanced AI is tested and deployed. - The proposal also links AI safety to sanctions, civic pressure and antitrust-style breakup remedies.

What happened: - One Planet – One People, a 501(c)(3) nonprofit, released the full text and enforcement architecture of its Global Protection Resolution on Sept. 29, 2026. - The organization says the resolution builds on its earlier launch of the One Planet Crisis Countdown Clock, or BRINK, and its call for a new UN Sustainable Development Goal on AI Education & Governance. - The release came during the final days of the UN General Assembly High-Level Week in New York. - Outgoing UN Secretary-General António Guterres warned this month against “a race to the bottom on AI safety,” and said AI needs “guardrails” that make it safe, transparent and accountable.

The details: - The resolution calls for a Global Frontier AI Safety Commission, or GFASC, with authority to test frontier AI systems before public release and block models that fail safety thresholds. - GFASC would be able to demand access to model weights and training documentation for audits. - The commission would have to use its own AI-assisted red-teaming and interpretability tools. - The proposal bars the commission from using audited companies’ proprietary systems or data for any purpose beyond the specific audit. - Funding would come from assessed contributions from frontier AI companies, scaled to global revenue and training compute, plus government funding. - The commission would be prohibited from holding equity stakes, taking consulting income or keeping financial relationships with companies it audits. - Staff would face a two-year cooling-off period before joining a company they audited. - Oversight would come from a rotating board with civil society, academic and government seats only. - The resolution says the commission would publish an annual public performance report on itself. - Mandatory human sign-off would be required before autonomous AI systems move funds above a defined threshold, alter production or safety infrastructure, deploy code to production, make targeting or use-of-force decisions, or materially affect legal rights, employment, credit or medical treatment. - The resolution sets a tiered incident ladder modeled on strict breach-notification rules in the EU's NIS2 directive and GDPR Article 33. - Critical incidents such as loss of containment, systems acting outside authorized scope, or theft of model weights would have to be reported within 24 hours. - Severe but contained incidents would require disclosure within 72 hours and a full report within 30 days. - A first confirmed failure would trigger civil fines based on global revenue and a mandatory 90-day remediation plan verified by an independent party. - A second failure within 24 months would trigger suspension of operating licenses, exclusion from government contracts and an independent compliance monitor at the company’s expense. - A third failure, or a single case of proven willful concealment of known catastrophic risk, would open formal proceedings before the Global AI Accountability Tribunal and activate the Karma Club civic response network. - A fourth-tier finding would pursue forced breakup or dissolution through existing antitrust and legal mechanisms, while preserving due process and appeal rights. - The proposed Global AI Accountability Tribunal would be a standing roster of vetted technologists, human-rights lawyers and ethicists. - The tribunal would be able to investigate confirmed complaints, subpoena documents and testimony from signatories, and issue public rulings binding on signatories. - For sanctions with legal force, the resolution calls for a Global AI Accountability Sanctions Mechanism modeled on the UN Security Council sanctions-committee process. - That sanctions mechanism would include an independent Ombudsperson and a due-process delisting and appeal channel. - The Karma Club network would activate only after a confirmed finding by the commission or tribunal. - The network would support consumer boycotts, public awareness campaigns, direct advocacy to lawmakers and regulators, institutional divestment petitions and shareholder-resolution support. - The One Planet Crisis Countdown Clock, branded BRINK, now reads four minutes and 30 seconds to midnight. - 1POP says the clock reflects loss-of-containment and deceptive-behavior incidents in frontier AI systems, the first confirmed fully autonomous lethal AI weapon strike causing civilian harm, concentrated wealth, underfunded safety efforts and the lack of binding international AI or autonomous-weapons agreements. - 1POP says it will update the clock monthly after review by its Board of Advisors. - The organization is calling on governments, AI companies, universities, companies, institutions, individuals and families to adopt its rules, use second-channel verification for unusual financial requests and treat AI-generated video and audio with skepticism. - Governments and others can read and sign the resolution and join the Karma Club at the resolution page. - Volunteers and interns can sign up at Benjamin DaVinci Institute.

Between the lines: - The proposal tries to fill a gap between soft international coordination and hard national enforcement. - Its structure mirrors existing global institutions, like the IAEA for funding and the UN Security Council for sanctions, to make the model feel more familiar to governments. - The Karma Club is meant to create immediate public pressure only after confirmed findings, which is designed to avoid rumor-driven campaigns. - By pairing pre-release oversight with after-the-fact penalties, the resolution pushes for a full regulatory stack rather than a single safety rule.

What's next: - 1POP says it wants governments to fund and empower the proposed commission and sanctions mechanism. - The group is also urging governments to pursue a binding international framework for frontier AI oversight and human control over lethal autonomous weapons. - The organization says it will continue publishing monthly BRINK readings and reasoning behind any changes. - 1POP is seeking sign-ons from governments, companies, institutions and individuals for the Global Protection Resolution and Karma Club.