Actions

High school athletic director accused of framing principal with AI-created racist rant

This is the first case involving artificial intelligence that the Baltimore County state's attorney has prosecuted.
mug1.jpg
Posted

Law enforcement in Baltimore arrested a 31-year-old man on Thursday morning, charging him with, among other things, “disrupting school activities.”

The suspect, identified in police documents as Dazhon Darien, is accused of faking the voice of the principal of Pikesville High School, where he worked, in an audio recording that attacked various members of the school and community.

During a press conference announcing the charges, Baltimore County Police Chief Robert McCullough said, “It’s been determined the recording was generated through the use of artificial intelligence technology.”

After the disturbing audio circulated on social media in January, officials in Baltimore launched an investigation. The principal, Eric Eiswert, has consistently denied that he made the comments, but he has been on leave since the incident occurred. Officials wrote in the charging documents that the audio clip “had profound repercussions” and “triggered a wave of hate-filled messages on social media and numerous calls to the school.”

In their investigation, law enforcement discovered that Darien, the school's athletic director, had researched and purchased an account with OpenAI. Investigators also had experts review the audio, and that analysis “indicated that the recording contained traces of AI-generated content with human editing after the fact, which added background noises for realism.”

One line from the fake audio recording attacked Black students at the school: “You know, I seriously don’t understand why I have to put up with these dumbasses here every day." The AI-generated audio also includes derogatory comments about Jewish people and teachers at the school.

When the story first broke, Scripps News took the audio to experts who analyzed and determined it to be authentic, however the union that represents the principal told us that their experts believed the audio was created by AI. Because of the conflicting accounts, we ultimately did not pursue our report.

A different expert told Scripps News on Thursday that the Baltimore case shows how anyone can use artificial intelligence.

“It's another feather in the cap of how easy it is to create AI-generated content,” says Mason Allen, the head of growth for the AI detection company Reality Defender. Allen described how some AI applications require just a few seconds of someone’s voice to generate a realistic reproduction – and a lot of this technology is free to use.

So, how do people figure out what’s real or not? Allen says, “It’s our stance that the average person shouldn't be the ones to have to wade through all of this and find out for themselves.” He added that social media platforms and generative platforms should step up to stop the spread of malicious AI content.

Baltimore County state’s attorney Scott Shellenberger revealed in the press conference that this was the first case that his office has prosecuted that deals with AI.

He told reporters, “As we dug into this a little bit more, it seems very clear to me that we may need to make our way down to Annapolis in the legislature next year to make some adaptions to bring the law up to date with the technology that was being used.”