Cisco is reported to have tested using artificial intelligence to draft security incident reports. According to accounts from multiple outlets, the approach can produce draft text for longer, structured incident write-ups, but the quality is not reliably consistent. The reporting highlights that generating usable long-form documentation with AI can be challenging even when the system is given detailed instructions and access to specific source material. One outlet notes that producing solid output may require extensive prompting, and that even with that effort the resulting reports can still contain errors and typos. Another outlet emphasizes that limitations around length, accuracy, and adherence to requirements affect the generated reports, indicating the method is not yet a dependable replacement for human work. Taken together, the sources describe a test where AI can assist in drafting portions of security incident documentation, but it still requires careful review and additional refinement to meet expected standards for correctness and clarity.