Insights

Editor’s note (24 September 2026): as of publication we have not located an official announcement of the measures described below; this article is based on industry briefings and is subject to official confirmation.

WIPO’s Madrid Registry says its Automated Correction System for irregularities, or MACS, has now gone live after technical integration with core European offices. Common formal defects, including loose goods and services wording and format conflicts, are expected to be identified earlier and blocked in real time. On one level, this is a speed story: if fewer routine defects travel downstream, average registration timelines may improve materially.

But that is not the part applicants should focus on first. Once automated checks move to the front end, much less room remains for broad drafting, soft-edged wording and the familiar habit of filing first and cleaning things up later. What used to be managed through later irregularity notices is starting to be controlled at the point of submission.

China’s revised Patent Examination Guidelines, effective from 1 January 2026, move legality review much closer to the centre of AI and big-data patent examination. CNIPA did two important things at once. It added an explicit Article 5(1) review standard for AI and big-data applications, and it also revised the examination baseline so that, where necessary, examiners may review the specification itself rather than looking only at the claims. That is a meaningful shift. Applications involving data collection, label management, rule setting or recommendation decisions are no longer judged only on whether the technical effect sounds persuasive. The file may also be read for obvious legal or ethical fault lines.

The official examples make the point in concrete terms. One concerns a facial-recognition marketing system that did not show lawful and compliant data acquisition. Another concerns an autonomous-driving emergency model trained to differentiate between people by age and sex. For applicants filing inventions around large-model training, corpus cleaning, data-labelling pipelines, alignment methods or vertical-model deployment, the signal is plain enough: training data and data-processing pathways are no longer a background black box the patent file can safely ignore. CNIPA has not published a standalone checklist of acceptable training-data provenance, but the distance between patent entitlement and data governance has clearly narrowed.

IP Australia’s current trade mark manual makes the point more plainly than before: a specification that is unrealistically broad can put intended use in issue at the examination stage. Filing an application will still usually be taken as indicating use or intended use, but that presumption no longer does all the work once the claimed goods or services are commercially implausible on their face.

That matters for applicants who have relied on wide, placeholder-style filings to reserve room first and define the business later. In the digital economy, the familiar combination of classes 9, 35 and 41 can bundle software, platform operations, advertising, training and content services into one application. Where the commercial story behind that bundle is thin, examiners now have clearer ground to demand a declaration of intended use or to press for a narrower specification at the front end.

Recent market commentary has started linking Singapore’s IPOS to a supposed AIGC-only green channel. Based on the current official record, that is not the safest way to describe what exists. What can be verified is narrower and more concrete. In 2019, IPOS launched AI², an AI-specific initiative that could take qualifying patent applications from filing to grant in as fast as six months. In 2025, Singapore shifted to SG Patents Fast, a broader programme designed to accelerate office actions rather than guarantee grant timing. Since 4 January 2026, new acceleration requests under SG Patents Fast have been suspended while IPOS reviews the programme.

That does not amount to a retreat from GenAI-related innovation. Read together, AI², SG Patents Fast, the 2026 suspension and IPOS’ wider positioning on generative AI and the creative economy point in a more nuanced direction. Singapore still values speed and commercial certainty in patent protection for emerging technologies, but it appears to be moving away from a narrow symbolic promise for selected sectors and toward more controllable, process-based acceleration tools that can sit alongside international cooperation routes. For GenAI toolmakers, content-tech platforms and creative software companies, the practical question is no longer whether there is a fresh AIGC-only lane today. It is how to align patent timing with fundraising, regional rollout and cross-border filing strategy.

Recent debate around the USPTO and Section 101 has made one point sound simpler than it really is: that an AI patent claim becomes easier to defend as soon as the application says the invention involves model training. The official materials tell a narrower and more useful story. From the 2024 AI subject matter eligibility update, to the August 2025 reminder memo, to the late-2025 MPEP change prompted by Ex parte Desjardins, the clearer direction is not a special AI shortcut. It is that claims framed as a black box that takes in data and produces a result remain vulnerable to abstraction, while claims that reflect a concrete improvement in how the model or system actually operates stand on firmer ground in the Section 101 analysis.

That distinction has real prosecution consequences. An application may contain pages of technical background, but if the claim is ultimately drafted as little more than “receive data, train or infer, output a result,” the applicant is still likely to face the familiar problem of an abstract idea implemented on a computer. What helps more is not a broad statement that the model is more accurate or more efficient, but a claim set that shows where the technical improvement lives and how it changes the internal operation of the system.

Debate around “shadow brand” filings in Latin America has moved up a gear. INAPI’s 2026 trademark guidelines further refine procedural practice, and opposition filings have already been folded into an electronic intake framework. The larger shift is not a cosmetic tweak to procedure. It is a change in what the office is willing to look at. The question is no longer confined to whether one sign resembles another. It is increasingly about how a filer behaves across a portfolio of applications.

What the public record clearly supports today is that INAPI is deepening digital procedure, clarifying opposition and invalidation routes, and continuing to treat bad-faith registrations as a problem that can be pursued over time. By contrast, the market narrative around a formal “filing metadata assessment” model, a trigger tied to more than 10 unrelated Nice classes, or a short deadline to produce a genuine business-use plan is better understood, for now, as a strong enforcement direction rather than a fully published rulebook. Even so, the signal is hard to miss: in Latin America, stockpiling filings without a credible commercial story is becoming much harder to defend.