An AI management system is a set of policies, processes and responsibilities through which a company governs its use of artificial intelligence. It begins with defining the scope – which activities, systems and locations are included – and continues with an AI usage policy that sets out what is and is not permitted within the company. It further includes a register of the systems used, a methodology for assessing risks and impacts, rules for data and its quality, requirements for suppliers, change management, an employee training plan, incident records and a mechanism for review by management. The most common mistake when implementing it is to create documentation that nobody uses. An auditor checks records from real operation, not the existence of directives, so the system must be connected to actual decisions on deploying tools.
See also: ISO/IEC 42001, Register of AI systems, Internal AI usage policy.
