Annotation

The article examines legal and ethical aspects of using artificial intelligence (AI) for genome editing, using the example of generating proteins used in genome editing using the CRISPR‑Cas method. It is shown that such proteins can be superior to natural ones, but they contain hundreds of mutations and insertions whose function is unknown, which creates unpredictable risks when they are used in medical practice. The author analyzes possible legal consequences: the liability of medical organizations to patients, the absence of recourse requirements for developers of AI products, the inadequacy of the existing regulation in Russia, and the need to strike a balance between innovation and safety. Measures to minimize risks are proposed, including independent testing, expanded informed consent, the creation of compensation funds, and data transparency.