On May 25, 2026, China's State Administration for Market Regulation (SAMR), led by Minister Luo Wen, and Russia's Federal Antimonopoly Service (FAS), headed by Maksim Shaskolsky, signed a bilateral antitrust cooperation memorandum covering 2026–2027. The text creates formal channels to exchange information on antitrust, unfair competition, and advertising matters, and — crucially — to coordinate on specific enforcement cases. Its declared sector scope is artificial intelligence, digital platforms, and semiconductors. It lands on top of two live SAMR probes against American chipmakers: a preliminary finding against Nvidia and a gun-jumping case against Qualcomm.
The strongest case for the deal
Begin with the steelman, because regulators have a real point. Antitrust is one of the few areas where international cooperation is genuinely uncontroversial in principle. Mergers and monopoly conduct routinely cross borders; a chip cartel or an abusive platform does not respect jurisdiction. That is why competition authorities have networked for decades — through the OECD Competition Committee, the International Competition Network, and dozens of bilateral memoranda. The United States and China themselves signed an antitrust cooperation memorandum years ago, and the EU coordinates case work with Washington as a matter of routine. Two large economies agreeing to share methodology and case intelligence is, on its face, ordinary good government. Information asymmetry is the enemy of sound enforcement, and pooling it can produce better-reasoned decisions and fewer conflicting remedies for the firms involved.
What makes this memorandum different
The problem is not cooperation as such; it is what these two specific agencies cooperate toward. The ICN and OECD frameworks rest on a shared lodestar — consumer welfare, measured by price, output, and innovation. Neither SAMR nor FAS operates under an independent, welfare-anchored mandate insulated from the state's strategic aims. Both are instruments of governments that are simultaneously parties to an industrial and geopolitical contest with the firms now under review. When the targets named in the same breath as the memorandum are Nvidia and Qualcomm — two American companies whose export-controlled products sit at the center of the US-China technology rivalry — the cooperation looks less like the ICN and more like coordinated leverage.
Look at the underlying cases. SAMR approved Nvidia's $6.9 billion acquisition of Mellanox in April 2020 subject to six binding behavioral conditions: supply on fair, reasonable and non-discriminatory terms, no forced bundling of GPUs with networking gear, no discrimination between customers, and maintenance of open-source commitments, among others, lasting six years. SAMR opened a compliance probe on December 9, 2024 — nearly five years later, and conspicuously after Washington tightened advanced-chip export controls — and on September 15, 2025 issued a preliminary finding that Nvidia had violated the Anti-Monopoly Law, with a deeper investigation to follow. Under China's 2022 AML revision, penalties can reach 10% of annual revenue.
The Qualcomm matter is cleaner on the law but louder on timing. On October 10, 2025, SAMR opened a gun-jumping investigation into Qualcomm's acquisition of Autotalks, an Israeli vehicle-to-everything chip designer. SAMR had told Qualcomm in writing on March 12, 2024 to file and not to close; Qualcomm said on March 14, 2024 it would abandon the deal, then completed it in June 2025 without notifying SAMR. A failure-to-notify case is legitimate on its own terms — companies should respect filing obligations. But a regulator that lets a matter sit for over a year and revives it amid a tariff standoff invites the inference that the enforcement calendar is set in the foreign ministry, not the competition bureau.
Why the procedural facts matter
Due process is the line between antitrust and economic statecraft. SAMR's own instincts can be sound: its November 18, 2025 draft guideline for internet platforms targets algorithmic collusion and "pick one of two" exclusivity — real harms — and frames itself as promoting innovation rather than imposing binding obligations. That is the proportionate, evidence-based posture good competition law demands. The Nvidia and Qualcomm cases are harder to read that way. A five-year-dormant conditions theory and a year-old notification lapse, surfaced precisely when each company is a bargaining chip in trade negotiations, strain the consumer-welfare rationale to breaking point.
Now add a second, sanctions-exposed regulator to the file. Coordinated enforcement means two agencies can open parallel cases against the same target, share evidence, and stack remedies and fines — multiplying legal exposure and settlement pressure without multiplying the underlying competitive harm. For globally operating firms, that is a meaningful escalation in regulatory risk that has little to do with whether consumers were actually harmed.
The cost of competition-as-coercion
None of this argues for less international antitrust cooperation. It argues for cooperation tethered to transparent, welfare-based standards and predictable process — the conditions under which firms can invest across borders with confidence. When enforcement becomes an extension of trade policy, the casualty is not just one American chipmaker; it is the predictability that the open, globally integrated technology economy runs on. Companies respond to politicized enforcement by fragmenting supply chains, pre-clearing every move, and pricing in geopolitical risk — frictions that fall hardest on innovation and, ultimately, on consumers in every market, including China's own. The SAMR-FAS memorandum may well be a routine document. The cases it will be used to coordinate are the reason to watch it closely.