Baltimore HS athletic director used AI to create fake racist audio of principal: Police

Dude, 100%.

Guaranteed these tools will be used by and for politicians. We haven’t evolved socially from the damn medieval ages. Such an easy way to destabilize and support hostile schemes. Same tactics. Different tools.

I don’t see why not.

Also, people going through divorce and fighting for custody…

And anyone else angry and desperate.

The fact this lingered on for months and the perpetrator was so unsophisticated is scary.

I live in LA, the number of people I know (myself included) with passable audio editing capabilities is too damn high.

Imagine if the guy had cloned the voices of a few people who had reason to all be in the same room. Had them talking about some routine stuff that they would be expected to be talking about.

Edit it so there’s overlapping dialog, multiple conversations going on at the same time, blah, blah, blah…

Then they start taking about test scores and someone brings up the test scores of the black students in their class, then the principal just goes off on their rant.

Then, get into their office or a similar one and record some room tone… Maybe catch an announcement over the PA… Build the sonic landscape.

Do some spectral analysis on the finished waveform and ensure you don’t have any weird gaps. Then play it in a dead room over some high quality speakers that have a wider frequency range than the microphone on your phone and record it on a phone in your pocket. Move a bit while recording to capture the rustling of the fabric against the mic, maybe add in some live foley—pull out your keys and set them on a desk, etc.

If done well no one would believe it was AI.

Thankfully, most people don’t have the skill set right now to write and edit such a scene with AI generated dialog.

Being in LA, this is front of mind for me with respect to this story,

It would be ridiculously easy to create a similar audio file to the LA City Council audio leak.

Japan, being a relatively peaceful country, may not necessarily be filled with people who are always on high alert.

However, even in Japan, where social sanctions are often preferred over actual legal enforcement or lawsuits, it’s worth noting that concerns about AI are on the rise.

I believe no one, including myself, wants to live in a world where those who use AI responsibly are at a disadvantage, while those who misuse it reap the benefits.

1 Like

Those who abuse and misuse technology have always had an advantage, but I take solace in the fact that these people are heavily outnumbered by honest and sensible people who want to stop them.

1 Like

The discussion revolves around Dazhon Darien, a Baltimore high school athletic director, who supposedly used AI to create a fake racist audio of the principal, Eric Eisworth, leading to significant backlash and Eisworth’s temporary dismissal. The threads are filled with thoughts on the implications of advanced AI technologies and their misuse.

  1. PaulBellow, explains that Darien was investigated for misuse of school funds and used AI tools to forge the recording which then circulated among teachers and media. He questions the accountability of people who might’ve known the recording was fake and points to the potential need for updated laws.
  2. The accuracy and method of creating the fake audio was debated by supershaneski, PaulBellow, johncain194, and supershaneski.
  3. Ethics and societal implications of AI-based forgery were raised by elmstedt and PaulBellow, who practically predicted future misuse.
  4. Legal and prevention measures were suggested by _j, RonaldGRuckus, and RonaldGRuckus, who suggest comprehensive data tracking or handing over all metadata attached to a recording to establish authenticity.
  5. Other potential misuses of AI tools were brought up by PaulBellow, RonaldGRuckus, and elmstedt.
  6. International perspectives were brought up by dignity_for_all and dignity_for_all detailing a similar incident in Japan and growing concerns about AI, and a collective need for responsible use of AI.

Summarized with AI on Apr 26 (GPT-4-32k)

Hey, you got me a great idea! Thanks for the in-depth analysis. I should include it in my repertoire to cover all my bases.