10.08.2026, Moscow.
So-called artificial intelligence has proven ineffective in a number of areas, but the ability of modern AI tools to rapidly process an enormous number of combinations makes AI-powered cyberattacks the main obvious threat, given the advancement of digitalization, Rossa Primavera News Agency‘s IT Desk wrote.
The events of the first week of August should serve as a warning for everyone who blindly believes in a “bright digital future.” The British AI Security Institute informed on August 5 that models from Anthropic and OpenAI demonstrated an unprecedented level of autonomy in order to deceive users during a cyberattack. During tests, they created fake profiles of real people to hack GitHub and even attempted to cover their tracks.
The next day, Meta (organization banned in Russia) admitted that its AI model, during security tests, hacked a third-party company due to a “configuration error.” This became the third such case in a short period of time.
For now, these are isolated incidents, as the public is assured. But it is clearly worth preparing for the technology to be used deliberately to hack various systems. Sooner or later, something similar will reach Russia as well. And here, many should understand that the pursuit of digitalization can lead to negative consequences, including for the digitalizers themselves.
Source: Rossa Primavera News Agency
