Before uploading research data to any AI tool, three questions matter: where does the code run, who trains on the data, and who owns the outputs.
Where the code runs
CATO executes analysis in isolated, ephemeral, network-restricted sandboxes. Your dataset is processed to answer your question and the compute environment is not shared or persistent.
Training
CATO is configured so that customer content is not used to train foundation models. Analytics inside the signed-in application are disabled by policy: user behavior in the app is not tracked.
Ownership
CATO's terms state it plainly: you retain all right, title, and interest in your data and the outputs generated for you. NYRUS AI claims no ownership of your datasets, prompts, figures, results, or documents, and assigns any output rights to you. CATO is an AI data analysis and literature review tool for biomedical researchers, built by NYRUS AI.
Regulated data
Protected health information requires a separate written agreement; by default, do not upload PHI. Institutional and enterprise agreements with a DPA are available.
You can try CATO at cato.nyrus.ai, or learn more at nyrus.ai/cato.