Mexico City calls for AI literacy for older adults: the digital divide is now the gateway for voice-cloning fraud

By CDMX Magacín · September 10, 2026.

Representative Pablo Trejo Pérez (District 15) presented a resolution to the plenary of the Mexico City Congress to protect older people's assets from the risks of artificial intelligence. The plenary approved it: it urges the Ministry of Education, Science, Technology and Innovation (SECTEI) to strengthen, through the community program PILARES and "in line with its budget availability", digital literacy in AI for older adults across the 16 boroughs.

The data behind the proposal is compelling. According to the 2025 National Survey on Availability and Use of Information Technologies in Households (ENDUTIH), 17 million people in Mexico do not use the internet, and for 9.2 million of them the main reason is a lack of knowledge. The age gap is the heart of the problem: 97.6% of young people aged 15 to 24 use the internet, compared with 57.8% of those aged 65 to 74, and only 30.3% among those over 75. Trejo Pérez put local figures to the problem by citing the case of Iztacalco, where more than 77,000 people over the age of 60 make up 19.3% of the borough's population.

The representative listed three types of AI-enhanced fraud already operating against this group: voice cloning for extortion, deepfakes simulating fake pension programs, and algorithmic impersonation to steal banking credentials. Citing data from the Cyber Police of the Ministry of Citizen Security (SSC), he noted that 33% of older people in the capital have been the victim of fraud or remote deception; and, citing the National Commission for the Protection and Defense of Users of Financial Services (CONDUSEF), that 63% of banking complaints from this group are linked to possible fraud or asset stripping.

The measure is worth reading with both feet on the ground: it is a call to action, not a law with earmarked funding, and its execution depends on SECTEI's budget availability and on PILARES's real capacity to reach the 16 boroughs with up-to-date content on the latest threats. The distance between passing a resolution and an older adult in Iztacalco receiving a hands-on session on how to spot a cloned voice remains, for now, an implementation unknown.

Our reading is that this case precisely illustrates the thesis we have been arguing about AI's short term: the harm does not come from a distant superintelligence, but from the industrialization of already familiar scams — phone fraud, banking phishing — to which voice cloning and generative deepfakes add a scale and a realism previously out of reach. The most exposed population is not the one that understands technology least in the abstract, but the one with the fewest practical tools to verify within seconds whether the voice of a "relative" or the notice of a "pension" is real. That is where digital literacy stops being a generic social policy and becomes a front-line cybersecurity measure.

This connects with something we have already seen elsewhere: what ends up being regulated is not the AI model's capability, but the vulnerability of the relationship that capability can exploit — Mexico City is trying to shield the bond of trust between an older adult and their bank, their doctor or their child, against a voice that no longer guarantees identity. It is a modest, local version of a problem that will recur in any city with an aging population and smartphones: the cheapest and most scalable defense is not banning the technology, it is teaching people to distrust the sensory evidence that used to be enough.

In the long run, the same AI that today clones voices for extortion is the one that within a few years will be able to detect those clones in real time, act as a personal assistant that screens suspicious calls, or explain in plain language why a message is fraudulent before the victim replies. The technological abundance we defend as a horizon does not exclude someone who is 75 today and uses the internet less: on the contrary, a society that distributes health, time and tools more equitably needs, first of all, for no one to be left out because they cannot read a screen. Initiatives like this one — modest, local, budget-dependent — are the unglamorous but necessary work of ensuring that this transition does not leave behind those who are already most vulnerable.