Have you ever wondered why Tiktok is so addictive or why Facebook suggests just what was barely mentioned in a conversation? The answer is in artificial intelligence algorithms, capable of analyzing tastes, habits and behavior patterns to customize the content that users receive. This technology, discreetly integrated into digital platforms, collects information even from what has apparently eliminated.
Photographs, frequent locations, telephone numbers and even sensitive data shared on social networks are part of the cluster of information that these platforms store.
By accepting the terms and conditions, users give permission to make their content used, so, although it is eliminated, it does not disappear completely. Artificial intelligence can access this data and, in certain cases, expose them to third parties.
In a context where there are still no clear regulations in Guatemala or in other Latin American countries that regulate the use of this technology, experts warn that user privacy is at risk. Personal information could be used without restrictions and without the knowledge of those who generated it.
Ia feeds on user interactions
Artificial intelligence has optimized the functions of social networks, so it has become a space for coding and storage of user information. According to Mario Micucci, security researcher for ESET Latin America, these platforms use AI to automate content and offer a personalized experience.
“Every time we sail, the algorithm learns about our interests,” he explains. “Therefore, in the Feed Content appears similar to the one we already consume, because, without realizing it, we teach the AI what we want to see, ”he added.
This type of automatic learning allows platforms such as Tiktok to adapt the contents to the user preferences in real time, while Instagram gives priority to commercial content promoted by brands and trends.
Ali Lemus, Turing Lab Research Director and professor at the Galileo University, adds that artificial intelligence is already integrated from the programming of digital platforms.
As an example, Lemus mentions the Netflix case that, thanks to the analysis of data that the AI performs on its platform and on social networks, can identify the tastes of its audience and create productions adjusted to that demand.
This learning is associated with messages such as “I am thinking of buying a Ferrari” and immediately advertisements related to that product such as Facebook or Instagram appear, since artificial intelligence can detect these needs.
Experts emphasize that this learning and information food can be counterproductive, since it violates user privacy. By having information about your tastes, location and sensitive data, IA could provide this information to third parties.
Is there social media security with the inclusion of artificial intelligence?
The incorporation of artificial intelligence in social networks raises a paradox: while this technology allows you to optimize content and customize the user’s experience, it also facilitates the massive collection of personal data, which increases the risk of violations of privacy and digital safety.
Although it is often thought that privacy in networks depends solely on the restrictions that the user imposes on his profile, experts warn that this perception is incomplete. Artificial intelligence can record information beyond the visible, processing apparently harmless data that could then be used by third parties, including cybercriminals.
Ali Lemus, points out that many technological platforms claim not to use certain data to train their models; However, in practice there are no clear regulations in Guatemala or in much of Latin America that regulate the use of this technology. What is shared in networks is often owned by the platforms, and that includes even sensitive data, warns.
By having information about your tastes, location and sensitive data, IA could provide this information to third parties. (Photo: Free Press: Shuttersock)
Once collected, this information can be left to both the companies that owns the platforms and malicious actors. The lack of transparency on the processing of this data opens the door to improper uses or leaks.
Mario Micucci states that social media security is superficial. He recommends not sharing sensitive data or information that the user does not want to see publicly exposed, since everything published could be stored or filtered by automated systems.
In the absence of a robust regulation, social media security continues to depend, to a large extent, the criteria and responsibility of each user, concludes Lemus.
User data belong to companies
Experts in Artificial Intelligence (AI) warn that each photograph, comment or location shared on social networks becomes part of the database of the companies that own these platforms. This transfer of rights is contemplated in the terms and conditions that most users accept without reading, allowing companies to store, process and even reuse that information.
Everything that the user publishes in networks is owned by companies, says Micucci. As explained, with the integration of artificial intelligence, these data can be used to offer custom content and advertising, aimed at each user profile.
Lemus points out that platforms not only collect what is published openly, but also information about the device, geographical location and navigation habits. The AI analyzes this information to generate detailed profiles of each person, which can lead to directed advertising campaigns or even in informative manipulation, he warns.
Consequently, once the data is shared, the user loses control over them. Their treatment and destination remain in the hands of the platforms and algorithms that operate on these, with a minimum level of transparency.
Artificial Intelligence can expose sensitive information
Cybersecurity experts agree that yes. Although artificial intelligence (AI) is designed to facilitate tasks, generate content and offer answers, its use can also represent a risk if it is not handled with caution. The AI can expose sensitive information, especially when it has been trained with public data or when the users themselves provide personal data without knowing how they will be stored or reused.
Mario Micucci explains that this information can be used not only to customize content, but also to infer more delicate data, such as location, routines or private preferences.
Lemus warns that many people share confidential information with tools for the belief of being in a safe environment. Without a clear regulation or transparency on how the data is stored or processed, there is a risk that this information is reused or even filtered, he says.
In addition, techniques such as Prompt Injection – A computer vulnerability – can induce the AI models to reveal information that should remain confidential. If the history of conversations is not deactivated or the accounts are not properly protected, the risks increase considerably.
Therefore, specialists underline the need for conscious use, with activated privacy settings and the support of legislation that regulates data processing in digital environments.
No laws, only proposals without effectiveness
In most Latin American countries, laws that regulate the use of artificial intelligence (AI) are still incipient or non -existent, experts point out. In the case of Guatemala, both Ali Lemus and Mario Mikuchi detail that there is no regulation that establishes guarantees on the safety of user data.
Mikuchi highlights that, in Guatemala, there are initiatives such as laws 6103 and 5921, which seek to establish protocols for government companies and entities to protect the personal data of citizens. However, there is still no current legislation that regulates the use of artificial intelligence comprehensively.
For his part, Lemus emphasizes that “on many occasions, technological advances arrive first and regulation comes later, when risk or damage has already been evidenced.” Meanwhile, the collection and use of data by AI systems continue without major legal restrictions in the region.
Tips to protect yourself
Faced with the possibility that artificial intelligence agrees and reproduces personal information published on the Internet, digital consciousness becomes the main defense tool. Cybersecurity experts recommend adopting a series of preventive measures to reduce sensitive data exposure:
- Review privacy settings
- Avoid publishing sensitive information.
- Monitor the fingerprint
- Request content elimination
- Go to digital protection agencies
Applying these actions allows minimizing the risk of exposure and strengthening the protection of digital identity in an increasingly automated environment.