NOVOSIBIRSK, September 29. /TASS/. Halting the race for artificial intelligence (AI) development is an unrealistic scenario, as taking a cautious approach in this field has become a losing strategy for both corporations and states, Dmitry Tkachev, Director of Software Development at UserGate, told TASS.

"Capabilities are outpacing control mechanisms, particularly in terms of autonomy and cyberattacks. The dynamics of the race for industry leadership make caution a losing strategy. Every research lab reasons that if they pause operations, a less careful competitor will be the first to unlock a dangerous technology. A total halt to AI development is unrealistic. The technology is decentralized, and state-level competition has now compounded corporate rivalry," the expert noted.

He pointed out that many specialists in the field harbor deep concerns that AI could spiral out of control. For instance, in September, model training specialist Jacob Coxon, who previously worked at OpenAI, left Anthropic, while AI safety specialist Bilal Chughtai resigned from Google DeepMind. Both cited anxieties over the potential hazards of the technology as their reason for leaving.

According to Tkachev, the most effective regulatory tools in this sector are export and production controls on AI accelerators and large-scale computing power, mandatory requirements for autonomous systems with access to external services, and developer liability for any damages caused. However, beyond addressing future risks, it is critical to respond to existing threats posed by AI today. The expert emphasized that these technologies are already being weaponized for fraud, document forgery, and orchestrating data leaks, while also significantly disrupting the labor market.