What people are actually arguing about.

This is contested now because Russia, China, and BRICS voices keep hammering that Western actions in Ukraine, the Middle East, and sanctions prove selective enforcement, while the US and Europe insist the order still constrains aggression. Recent multipolar push via SCO, BRICS expansion, and Lavrov speeches has sharpened the split as institutions like the UNSC stay deadlocked.

Closed labs and some policymakers push restrictions or unequal treatment of open weights amid national-security and misuse fears, while open advocates and some big tech warn this creates moats that protect incumbents' API rents. Recent US frameworks and NTIA reports, plus media waves and lab lobbying, have sharpened the fight as open models rapidly close the capability gap.

Frontier labs are negotiating federal pre-release reviews and capability-based rules while publicly backing pacing; open advocates say the real target is competitors and Chinese open models. Recent bill language and Frontier Model Forum coordination plus Amodei/Altman statements on slowing the frontier have made the capture charge the default counter-framing this month.

Russia's war in Ukraine and conflicts like Gaza have intensified claims that the West enforces rules only when convenient, while multipolar advocates (BRICS, Global South) push UN Charter sovereign equality instead. Recent multipolarity talk from Lavrov and others, plus critiques of selective enforcement, keep this boiling as US/NATO influence is seen as declining.

Labs and policymakers are pushing capability thresholds that would force licensing or bans on open weights while closed frontier APIs get lighter or negotiated rules; recent RSP proposals, bill language, and government pre-release review frameworks make the split concrete. Open advocates see this as big-lab moat-building timed to Chinese open models and commercial OSS use; safety/national-security voices treat uncontrolled weights as the vector that makes pacing or containment impossible.

Labs and safety advocates are pushing pre-release audits and export-style controls that would hit open releases hardest, while open-source backers see this as the start of a cartel locking down intelligence. Recent lab working groups, Amodei’s block-power calls, and Chinese open models catching up have made the stakes immediate: open weights are now good enough for real apps but still lag the absolute frontier.

This is contested now because Russia's war, Gaza fallout, and rising BRICS/SCO activity have exposed enforcement gaps; Western leaders still invoke the order while Lavrov and Global South voices call it hypocritical and push UN Charter 'sovereign equality' as the real foundation. Recent multipolarity talk and US alliance strains make the clash sharper than abstract IR theory.

This is contested now because Russia's war in Ukraine and rising BRICS/SCO activity have made multipolarity rhetoric mainstream, while Gaza and selective enforcement expose double standards. Recent Lavrov speeches and Global South pushback frame the UN Charter as already containing the fix the West never applied.

Labs (Anthropic, OpenAI, Google) have been meeting on an industry-led auditor after a stalled White House EO and Hassabis's FINRA proposal; Amodei wants blocking power while critics call it self-serving barriers that freeze out startups and open models. The fight sharpened as CEOs publicly floated coordination just as open-weight alternatives and Chinese models closed the gap.

This is contested now because of ongoing wars in Ukraine and Gaza, plus rising multipolar rhetoric from Russia, China and the Global South, where Western invocations of the order clash with accusations of selective enforcement and exemptions for allies. Recent shifts include explicit multipolar declarations by Lavrov and others, plus critiques that US actions undermine the very institutions it created.

The fight is hot because Russia, China, and Global South voices keep hammering that the West never practiced sovereign equality while the US and allies insist the postwar order (UN + alliances) is what keeps the peace. Recent multipolar summits, BRICS expansion talk, and continued Ukraine fighting have sharpened the split over whether the system is collapsing or being defended.

The argument is white-hot because the war continues, long-range strikes into Russia are debated, and NATO's open-door policy collides with Russian red-line rhetoric years after the full-scale invasion. Recent talk of security guarantees, Finland/Sweden accession, and 'indivisible security' keeps both camps citing the same OSCE/UN texts against each other.

The fight is hotter because open-weight releases (and near-frontier Chinese models) keep closing the gap on closed APIs while US labs and safety advocates push pre-release reviews that may treat open and closed differently. Recent US framework talks and lab lobbying have made the split concrete: one side sees open weights as democratizing and inspectable, the other as irreversible proliferation of cyber/bio/autonomy capabilities.

Reports of a nearing US framework requiring frontier labs to give agencies like the NSA up to 30 days access for national-security checks, plus proposals from Demis Hassabis for a Frontier AI Standards Body and Bessent for a FINRA-style regulator, have ignited the fight. Labs are already negotiating the rules; the definition of 'frontier model' and treatment of open vs closed weights is unresolved, raising stakes over whether this cements state control or is essential risk mitigation.