Russian Startup Raises $0.12m to Defend AI Systems From Hacks and Leaks

As artificial intelligence spreads across industries, one Moscow-born project bets that its real-time defense system can outpace hackers and rogue prompts.

A Moscow-based startup, backed by the venture studio of the Moscow Institute of Physics and Technology (MIPT), has secured 10 million rubles (about $0.12m) to build HiveTrace, a cybersecurity platform designed to protect AI systems from breaches, leaks, and malicious use.

HiveTrace monitors large language models in real time, instantly spotting and blocking suspicious activity. The system is engineered to neutralize so-called prompt attacks, prevent data leaks, and stop unethical behavior by AI agents before it spreads.

The timing is deliberate. AI adoption in Russia is accelerating fast: about 70 percent of mid-size and large companies already deploy AI tools in daily operations. But that growth has triggered a surge in cyberattacks, raising alarms about how vulnerable next-gen systems really are.

HiveTrace is already running pilot projects with financial institutions, telecom providers, industrial companies, and a wave of generative AI startups. The company’s founders believe demand for reliable AI defense will extend beyond Russia, especially as enterprises worldwide brace for the risks tied to rapid AI rollout.