Does militarization digitize?: The ANGEL plan can violate human rights

We already share data to companies, what is the difference in this case?

Erik Huesca, a physicist from UNAM and a PhD in AI from the University of California at Berkeley, commented that currently practically all people connected to the Internet are constantly monitored by private companies, but stressed that a system of these characteristics can be problematic. due to the ethical biases with which it is programmed.

There is a risk that a system of “artificial brutishness”—as he sarcastically calls them—programmed with the information that tattooed people were in jail, signals that any individual with tattoos could be a potential criminal.

Garcia shared Dr. Huesca’s concerns that private companies also have large amounts of user data, but noted the fact that a government would collect such amounts of information for surveillance purposes, including biometric data, is particularly troubling.

The specialist said that users are not required to use technology services if they do not wish to provide their data, however, with a proposal like the ANGEL plan, the measure is “unavoidable” for everyone.

“Both the use of data by companies and governments are quite problematic and neither party justifies a deployment of mass surveillance systems,” he commented.

The experts also agreed that there are no regulatory frameworks in Mexico with which to prevent this type of technology from being used abusively against citizens. In this sense, García cited the Pegasus case, through which the government spied on activists and journalists with impunity.

“If they were not able to prevent, protect and punish when there were cases of surveillance focused on a small number of people, much less will they be able to do it when the surveillance is generalized,” he said.

Data centralization, the latent risk of hacking

To make this type of technology work, it is necessary to have a large database. In Mexico, the attempt to create such a large information repository was called the National Register of Telephone Users (Panaut), which was declared unconstitutional by the Court of Justice.

Huesca said that a system with these characteristics can be the object of cyber attacks or leaks to obtain all the available information, which would be quite harmful for citizens, since malicious agents would not only obtain their personal information, but also biometric data that represents a level much higher vulnerability.

On the other hand, García stressed that the context of collusion and corruption that usually exists in public security corporations is also an argument for not trusting Marcelo Ebrard’s plan.

Although in Mexico there are specialists in many areas of technology and Artificial Intelligence, Aldo Luévano, CEO of Roomie IT, stressed that in this initiative, if it is carried out, hyper-specialized experts capable of executing a detailed curatorship will also be needed. on the assumptions made by the AI, with the objective of minimizing the possible errors of the tool.

Miguel Hernández y López, engineering manager at the cybersecurity firm CheckPoint, added that more preparation is also required in the management of technological systems to know the risks and generate action plans when finding a type of incident.

On the other hand, Dr. Huesca, who was president of the Mexican Academy of Informatics, concluded that any AI system that is applied on a massive scale must make transparent the ethics with which it was made, that is, what were the values ​​with who was programmed to analyze data, as well as who was responsible for creating the technology and what logic is behind its operation.