WIPO Classifies Artificial Intelligence as an “Out of Reach” Capability for Mexico

What the Outlook measures, and how it departs from the Global Innovation Index

The Innovation Capabilities Outlook 2026, presented by WIPO in Geneva on January 29, 2026—under Director General Daren Tang and Deputy Director General Marco Alemán, with oversight by Chief Economist Carsten Fink and by Hausmann, and a team led by Julio Raffo and Muhammed A. Yildirim—does not count innovation inputs and outputs the way the Global Innovation Index does. It measures an economy's repertoire of capabilities—how many fields it commands and how complex they are—and the proximity between those fields, that is, how close a country stands to moving into a new field from what it already knows how to do.

It analyzes 2,508 fields grouped into four dimensions. Science is measured through scientific publications indexed in Scopus via the OpenAlex database, taking the top ten percent most cited. Technology is measured through international families of granted patents that sought protection outside the country of origin, combining WIPO databases with the European Patent Office's PATSTAT and classified by four-digit IPC code. Entrepreneurship is measured through international trademark records from the Global Brand Database, clustered algorithmically rather than by Nice Classification alone. Production is measured through manufacturing exports from the United Nations COMTRADE database.

The Global Innovation Index—which in its 2024 edition placed Mexico 56th among 133 economies and in 2025 placed it 58th among 139—asks whether a country has the ingredients of innovation and whether it converts them into results. For technology policy purposes, the Outlook reveals something equally important: given what a country already commands, which new fields are within its reach and which are not? The answer is organized into four categories—established, at risk, within reach, and out of reach—resting on two axes the report's technical notes do document: feasibility, understood as the closeness or relatedness of a field to the capabilities the country already holds, and attractiveness, understood as the field's complexity, which is where the report itself locates the rewards when it notes that “relatedness indicates the probability that an economy will master a given field, and complexity represents potential rewards.”

Mexico's portrait: high diversity, low complexity

The portrait the Navigator draws of Mexico is summarized, in the instrument's own words, as “a large innovation ecosystem with high diversity of low complex capabilities.” Translated:

Mexico does many things, but few difficult ones.

In the information and communication technologies domain, which groups 32 fields, Mexico has zero established capabilities and zero within reach; all 32 are out of reach, and among them is artificial intelligence, whose complexity rank sits at roughly 1,260 on the global scale. In semiconductors and optics, 16 fields, the picture is identical: zero and zero, all 16 out of reach. In biopharma, 28 fields, Mexico holds a single established capability, and it is food preservation.

Mexico assembles, but Mexico does not invent

Where the country does accumulate capabilities, they confirm its nature. In machinery and transport, a production domain with 160 fields, Mexico holds 54 established capabilities, all of low complexity: passenger cars, auto parts, trucks, electric motors, transformers, cable and wire. In food, beverages, and tobacco—a domain the instrument classifies as entrepreneurial because it measures it through trademarks—there are 23 established capabilities, again of very low complexity. The pattern is unmistakable: Mexico assembles, but Mexico does not invent. Characterizing Latin America and the Caribbean, the report describes a system driven by science and production that, in its own words, “struggle[s] to translate these into entrepreneurial ventures and technological breakthroughs,” and closes with a comparison that stings for its simplicity:

“Japan demonstrates five times Mexico's diversity despite a similar demographic scale.”

The margin lies in diversity, not in volume

In my previous piece I proposed multiplying resident patents fivefold. The Navigator suggests that this is not the principal lever. Mexico's relevance potential—how much the country weighs in the global volume of innovation—barely rises from a current 0.79%, up from 0.76% in the past and placing the country 24th, to a potential 0.82%: an almost negligible margin. Diversity potential, by contrast, nearly doubles: from a current 12.1%, ranked 59th, to a potential 23%. Put differently, Mexico's room for improvement lies not in producing more of the same but in widening its repertoire of capabilities toward adjacent fields it does not yet occupy.

Sophistication tells a similar and troubling story: 43.4% today, ranked 54th, against 44% in the past. Both diversity and sophistication moved backward over ten years. The pathway the instrument assigns to Mexico, “Breaking Out”—glossed as “deep expertise is opening doors to new complex fields”—describes precisely that wager: Mexico's route runs through diversifying into the adjacent, not through thickening the volume of a handful of fields.

None of this makes multiplying resident patents a mistake. It makes it an outcome indicator rather than a strategy, and pursuing it without first widening the capability repertoire mistakes the effect for the cause.

What the Navigator does not measure

One caveat follows from the instrument itself. That artificial intelligence appears as “out of reach” does not mean software is not written in Mexico, nor that there are no brilliant Mexican engineers in the field; it means Mexico does not show revealed comparative advantage in international patent families in that field, which is what the instrument measures. And Mexico's density in the entrepreneurial dimension speaks neither to its gross domestic product nor to its employment, but to trademark records in the Global Brand Database. The Navigator measures the repertoire and proximity of technological capabilities, not welfare or aggregate economic activity. Conflating the two would be exactly the error the report seeks to avoid—the report that finds only ten percent of economies meet their technological potential and estimates that the world underdelivers some 339,000 technological innovations a year, twenty-six percent of the total, according to its Table 3.4.

What this means for Mexico's 2026 industrial property reform

What all of this means for the April 2026 reform of Mexico's Federal Law for the Protection of Industrial Property (Ley Federal de Protección a la Propiedad Industrial, LFPPI) is the part that matters most to those of us who litigate and advise in this field. The statutory decision deadlines—five months for trademarks, one year for patents from the start of substantive examination—the nearly five hundred new examiner positions that Economy Secretary Marcelo Ebrard announced on June 1, 2026, when he appointed Vidal Llerenas Morales to head IMPI in succession to Santiago Nieto, and the Specialized Technical Committee for mandatory decisions are all improvements to the pipeline. They speed the flow of files. They do not create a single technological capability.

I already argued, in analyzing that Committee's Guidelines, that a body without a substantive standard will struggle to clear the backlog. The Navigator adds a deeper layer to that criticism: even if IMPI decided every file in record time, it would still be processing, above all, foreign patents and low-complexity trademarks, because that is what the country's repertoire produces. Modernizing the registration window of an economy that does not command complex technologies is indispensable, but it should not be mistaken for innovation policy.