Insights

The USPTO has confirmed that, starting July 18, 2026, new U.S. trademark applications filed under Sections 1 and 44 of the Trademark Act will receive eight-digit serial numbers beginning with “50.” On paper, it looks like an administrative reset because the current “99” series is nearing exhaustion. In practice, it is the kind of small back-end change that can expose very old assumptions inside law-firm and in-house trademark systems.

The agency also made clear what does not change. Priority will still be determined by filing date and time, not by the size of the serial number; examination timing is not supposed to shift; and Madrid extensions of protection to the United States will continue to begin with “79.” The real problem sits elsewhere: countless docketing rules, spreadsheet macros, reports, and data pulls quietly assume that newer U.S. applications always carry larger numbers. From July 18 onward, that assumption stops working.

After hosting the 19th IP5 Heads of Office Meeting on 12 June, the Japan Patent Office said on 16 June that the five offices had agreed on new directions for AI-focused co-operation and would establish a dedicated working group to carry the discussion forward. Read literally, that announcement is still framework-level. It does not yet amount to a single binding AI patent rulebook shared word for word across Japan, Europe, China, Korea and the United States.

What matters for applicants is something narrower and more practical. The JPO already has updated AI case examples, refreshed handbook materials, an expanded internal team supporting AI examinations and external AI Advisors. Put those pieces next to the IP5 announcement and the likely direction becomes easier to read: AI prosecution is moving away from broad functional drafting and toward closer questioning on technical route, disclosure density, technical effect and whether the claimed result can really be tied to reproducible implementation choices.

China’s revised Trademark Law was formally adopted on 26 June 2026 and will take effect on 1 January 2027. The first wave of official explanations released in early July makes one point hard to miss: this is not a narrow procedural update. It is a broader reset of how trademark filing, use, administration and enforcement are expected to fit together.

The headline issue is tougher action against bad-faith filings and stockpiling. That matters. But the more consequential shift is structural. The law is moving away from a system that could still tolerate “file first, sort it out later” behaviour and toward one that asks a more basic question from the start: why is this mark being filed, how will it be used, and does the rights strategy still make sense once market order and public impact are taken seriously?

South Korea’s latest IP moves are easier to understand together than apart. On 25 June, MOIP outlined its IP-AX coordination framework, signalling that AI-assisted invention, AI-enabled examination and related policy adjustments will be handled in a more integrated way. On 3 July, MOIP then publicised a four-party cooperation arrangement with MFDS, the Korea Customs Service and the Korea Cosmetic Association aimed at tightening the response to counterfeit cosmetics moving through cross-border channels and online platforms.

Read separately, one story looks like patent governance and the other looks like brand enforcement. Read together, the picture is sharper: South Korea is trying to place examination standards, platform governance and border enforcement inside a more connected regulatory model. For applicants, brand owners, cross-border sellers and platform operators, that is not a symbolic shift. It is a practical one.

The USPTO has formally finalized a rule that will make late corrective petitions in patent matters harder once the delay runs past one year. Published in the Federal Register on June 24, 2026 and effective August 13, 2026, the rule lowers the point at which the Office will require a fuller factual showing from two years to one. It reaches more than abandoned applications: delayed priority or benefit claims, unintentionally delayed maintenance fee payments, and missed time limits in Hague international design matters are all in scope.

This is not just a procedural tweak for outlier cases. It changes how applicants and patentees should think about internal monitoring, escalation, and rescue timing. After August 13, petitions filed more than one year after the missed action will face a more demanding information burden and the higher petition-fee tier. A slow internal review process will become harder to defend.

Canada’s patentability analysis for AI and computer-implemented inventions is moving into a markedly less forgiving phase. CIPO’s March 2026 Practice Notice put the spotlight squarely on the “physicality” requirement implicit in subsection 27(8) and section 2 of the Patent Act: if a claim merely has a computer process an abstract algorithm in a well-known way, without a discernible physical effect, physical change, or an improvement in the functioning of the computer itself, eligibility risk rises quickly.

By late June, practitioner feedback suggests that this is no longer just a policy paper people cite in seminars. Examiners appear to be applying the framework more consistently in day-to-day prosecution, especially for AI models, software rulesets and data-driven diagnostic methods. Saying that a model predicts better, ranks better or diagnoses better is no longer enough on its own. Applicants are increasingly being pushed to show where the physicality lies and why the claimed invention is more than a bare practical application of an abstract idea.