02.06.2026 17:25:00
Дата публикации
Artificial intelligence is becoming deeply integrated into everyday life — from entertainment recommendations to banking and credit decisions. Yet with convenience comes growing concern about digital rights and privacy.
In an interview with Youth.kz, Elzhan Kabyshev, Head of Legal Practice at the Eurasian Digital Foundation, explained that personal data has become the key resource of the AI era. Every day, users leave behind digital traces: purchases, photos, geolocation, and financial transactions. These datasets form the foundation for training neural networks.
Kabyshev noted that Kazakhstan’s Law “On Artificial Intelligence” directly classifies AI systems as objects of informatization. Any interaction with such systems involves data exchange, including personal information.
AI can be a powerful tool for security: detecting fraud, blocking malware, and preventing cyberattacks. However, the same technologies are exploited by malicious actors to produce deepfakes, fake websites, and convincing phishing emails. This duality makes AI both a safeguard and a source of new vulnerabilities.
The expert stressed that banning technology is not the solution. Instead, resilience must be built through three elements: digital literacy, critical thinking, and systematic awareness of threats. Users must understand how their data is collected and processed, and demand transparency from companies.
Transparency is crucial because ecosystems often involve dozens of services exchanging information. Users see only the outcome — targeted ads or recommendations — without knowing how the system reached its conclusions. Companies are therefore obliged to explain processing principles and provide tools for user control.
Equally important is the right to appeal. Algorithmic errors can block access to services or generate inaccurate ratings. Kabyshev highlighted one of the most serious risks: “model inversion,” where training data can be partially reconstructed by attackers through targeted queries. This means that even anonymized datasets may be vulnerable.
Thus, AI simultaneously strengthens protection and creates new attack vectors. Preserving human oversight over critical decisions and developing legal mechanisms for appeals are essential to safeguard digital rights in the age of artificial intelligence.
Kabyshev noted that Kazakhstan’s Law “On Artificial Intelligence” directly classifies AI systems as objects of informatization. Any interaction with such systems involves data exchange, including personal information.
AI can be a powerful tool for security: detecting fraud, blocking malware, and preventing cyberattacks. However, the same technologies are exploited by malicious actors to produce deepfakes, fake websites, and convincing phishing emails. This duality makes AI both a safeguard and a source of new vulnerabilities.
The expert stressed that banning technology is not the solution. Instead, resilience must be built through three elements: digital literacy, critical thinking, and systematic awareness of threats. Users must understand how their data is collected and processed, and demand transparency from companies.
Transparency is crucial because ecosystems often involve dozens of services exchanging information. Users see only the outcome — targeted ads or recommendations — without knowing how the system reached its conclusions. Companies are therefore obliged to explain processing principles and provide tools for user control.
Equally important is the right to appeal. Algorithmic errors can block access to services or generate inaccurate ratings. Kabyshev highlighted one of the most serious risks: “model inversion,” where training data can be partially reconstructed by attackers through targeted queries. This means that even anonymized datasets may be vulnerable.
Thus, AI simultaneously strengthens protection and creates new attack vectors. Preserving human oversight over critical decisions and developing legal mechanisms for appeals are essential to safeguard digital rights in the age of artificial intelligence.
