RTK-Information Security, a subsidiary of Rostelecom operating under the Solar brand, has introduced strict clauses regarding artificial intelligence in its new contractor agreements. The company now mandates that any use of AI tools during project execution must be explicitly approved by the client in advance.
Violations of this policy carry significant financial and legal consequences. Solar reserves the right to reduce the contract price by 30%, terminate the agreement unilaterally without compensating for losses, or demand that the work be redone from scratch without the use of AI tools.
The new requirements stipulate that if AI is used with permission, contractors must manually verify all outputs, describe the specific tools and scenarios applied, and ensure that no sensitive data is used for model training. Furthermore, contractors bear full responsibility for any intellectual property disputes arising from AI-generated content.
The company justifies these measures by citing risks related to data leakage, potential errors in AI-generated code, and the lack of transparency regarding how external AI models handle processed information. Solar emphasizes that these rules are particularly critical for projects involving state infrastructure and sensitive data.
While Russian law currently lacks specific regulations on AI usage in contracts, the Ministry of Digital Development has proposed a bill requiring disclosure of AI involvement, expected to take effect in September 2027. Experts note that while the trend of restricting AI is growing, proving its unauthorized use remains technically challenging.
