model training
Latest stance: training permitted
“take any action that would cause any portion of the Services to be placed in the public domain; (n) remove, alter, or obscure any proprietary notices of Semrush, its licensors, or its suppliers included in the Services; (o) use the Services to create, send, or store infringing, obscene, threatening, libelous, or otherwise unlawful or tortious material, including material that is harmful to children, and including spam, unsolicited commercial, or other harmful emails, or otherwise harass, threaten, demean, embarrass, bully, or harm any individual; (p) harvest or scrape any content of the Services or any related system or service; (q) perform any fraudulent activity including impersonating any person or entity, claiming a false affiliation or identity, or accessing any other Account without permission; or (r) use, access, copy, or process any Semrush intellectual property (including insights, analyses, suggestions, graphs and other outputs from the Services, including generative AI outputs) or Semrush Confidential Information as inputs/prompts into, or for the purposes of developing, training, improving, fine-tuning, testing, or otherwise enhancing, any artificial intelligence, machine learning, or similar technologies (including but not limited to large language models, generative AI systems, predictive analytics, or algorithmic models), including by means of combining or commingling the foregoing with other datasets for such purpose (except for Semrush’s official integrations embedded into an AI system).”Open citation
