Artificial intelligence tools or models that have been introduced to a system, such as a network, without the knowledge or approval of the organization's IT, cybersecurity, or data governance teams, thus creating the potential for vulnerabilities and risks to be introduced to the larger system
