Mayor’s office seeks ethics probe after IG shares racist AI video

Baltimore’s mayor’s office has asked the city ethics system to examine whether Inspector General Isabel Cumming crossed a line when she shared a racially charged AI video targeting Mayor Brandon Scott, a move that has sharpened scrutiny of the city’s top watchdog at a moment of already frayed trust. Cumming apologized after deleting the post, which showed Scott with a cigar, a drink that appeared to be alcohol, luxury shopping bags and a suitcase overflowing with cash.

The controversy matters because Baltimore’s ethics law applies to all city officials and employees, and it is built around broad standards of fairness, impartiality and restricted conduct rather than any rule written specifically for artificial intelligence. The law bars a public servant from intentionally using the prestige of office for private gain, and it also forbids the misuse of confidential information. The Ethics Board, meanwhile, is authorized to investigate complaints, determine whether they fall within its jurisdiction and, if warranted, issue sanctions.

That framework gives J.D. Merrill, Scott’s chief of staff, a formal lane to press the issue. His letter to the Ethics Board called Cumming’s conduct “deeply inappropriate, misleading, damaging and racist,” and asked for a review of what corrective action, if any, is warranted to preserve accountability, fairness and objectivity inside the inspector general’s office. The city also remains locked in a separate legal fight with Cumming over access to records tied to the Mayor’s Office of Neighborhood Safety and Engagement, adding another layer to a dispute that is already testing institutional norms at City Hall.

Cumming has previously won Ethics Board guidance protecting her right to speak on personal social media accounts, when the board said last fall that her posts about an AFSCME Local 44 election did not violate the city’s “prestige of office” rule. That history makes this case more than a single bad post. The question now is whether sharing an AI-generated racist caricature of the mayor undermines the credibility of future investigations by the official charged with rooting out waste, fraud and abuse in Baltimore government. If the public cannot trust the watchdog to stay above partisan and racialized attacks, every finding that follows will face a harder climb.