BALTIMORE, MD (Wednesday, March 20, 2024) – Today, Mayor Brandon M. Scott issued an executive order designed to govern the use and development of artificial intelligence (AI) within city operations. This proactive step intends to ensure responsible use of generative AI tools as they become more prolific and offer additional opportunities to support government services. This executive order reflects Baltimore's commitment to innovation, accountability, and the welfare of its residents.

"City government has routinely played catch-up in terms of technology, but this administration is trying to change that. We know that artificial intelligence is moving incredibly fast, presenting both unprecedented challenges and opportunities for city services," said Mayor Brandon M. Scott. "We must proceed with caution and the utmost responsibility to ensure the best interests of Baltimore’s residents are centered as the use of AI becomes increasingly common. This executive order lays the foundation for a future where the use of AI technologies is guided in a manner that is transparent, secure, and beneficial for all Baltimoreans."

The executive order on the use of generative AI for city employees does four main things:

• Outlines the city’s general position on using generative artificial intelligence tools
• Explains guiding principles drawn from guidance already outlined by the White House and National Institute of Standards and Technology (NIST)
• Provides specific guidance on usage for city employees
• Outlines next steps, including the establishment of an advisory committee on artificial intelligence.

In the coming weeks, the City will establish an advisory group whose immediate priorities include assessing the current use of generative artificial intelligence technologies across city agencies, researching and supporting policy development, offering education and training for city employees, and developing risk profiles for safe and productive use cases.

“We are grateful to Mayor Scott for his leadership and for the guidance his office has issued to employees,” said the City of Baltimore’s CIO Todd Carter. “As we work to put these principles into action, the Baltimore City Office of Information and Technology will continue to be a resource to employees who would like to learn about these new AI tools and to help employees use generative AI responsibly and effectively.”

AI technologies, particularly generative AI, represent rapidly-evolving fields that offer significant opportunities that could: advance our understanding of complex phenomenon, predict or suggest responses and interventions and, improve government services. However, its use also presents many still unresolved ethical, security, and policy risks. Given the pace of development, potential usefulness, and inherent risks and ethical considerations, this executive order aims to provide guidance to employees using these tools, as well as establish the governance structures necessary for continued analysis and policy development over time.

The executive order is grounded in core principles outlined therein, which align with guidance from the White House and the National Institute of Standards and Technology (NIST). Additional key elements in the executive order include:

• Establishment of an advisory group comprising city agencies and external stakeholders to oversee AI implementation and compliance
• Restriction on technologies that could fake or imitate individuals' appearances or voices
• Disclosure of use of generative artificial intelligence for deliverables used in decision making where the product uses substantial outputs produced by a generative artificial intelligence system
• Separation of city-related work from personal work when using generative artificial intelligence tools not managed by Baltimore City