🧠 The Real Status of AI Agent Adoption Revealed

Russian businesses are overwhelmingly leveraging point AI solutions (86% of companies), yet full-scale integration of artificial intelligence into core processes still lags far behind public success stories.

These are the findings of experts from the consulting firm Technology of Trust (TeDo) in their study, “AI Agents: Current Realities and Prospects in Russian Business.”

The primary obstacle to adopting autonomous AI agents remains business process unreadiness and management resistance. According to the TeDo survey, only 13% of domestic business leaders are willing to relinquish control over neural network decisions entirely, while the vast majority remain wary of full algorithmic autonomy. The situation is further compounded by a lack of funding: most enterprises allocate no more than 3% of their total investment budget to AI.

Among the key conclusions of the study, experts highlight several defining trends:

  • A Shift in the Human Role: Instead of crafting complex prompts, humans are transitioning to defining intents and goals, leaving the autonomous AI agent to independently formulate plans and select tools.
  • The Platform Approach: Point-solution chatbots have reached the limits of their effectiveness. The future belongs to comprehensive AI agent platforms that establish a unified framework for managing hybrid human-machine teams.
  • GPU Shortages and the Cloud Boom: Hardware constraints in the GPU market have made cloud computing rentals the key driver of growth—this sector reached 17 billion rubles in 2025, with experts anticipating exponential expansion ahead.
  • Synthetic Data: As human-generated datasets approach exhaustion, AI providers are shifting to large-scale synthetic data generation to train their algorithms.
  • The Struggle for Data Governance: Cybersecurity risks (a concern for 30% of companies) and poor corporate data quality could lead to the freezing of up to 60% of immature AI projects.

The authors of the report emphasize that the future evolution of this technology will demand a massive increase in compute capacity—by up to a factor of 10,000—alongside a shift in government AI regulation from voluntary ethical guidelines to mandatory regulatory requirements and compliance audits.

Source: Telecom Daily