Weekend Reading and Listening

ASP Teach-In with Barb McQuade

This week, I had a sobering conversation about the use of disinformation as part of the authoritarian toolkit with

Barb McQuade

, former United States Attorney for the Eastern District of Michigan from 2010 to 2017 and author of Attack from Within. You can watch our conversation above (and if you like my t-shirt, you can order one here). This conversation was sponsored by my non-profit,

The American Sunlight Project

. Every month, we host teach-ins with folks in and around the counter-disinformation field. Thank you

Matthew Facciani

,

Paul kirkpatrick

,

Janet L Stevens

,

BCz

,

Amy Cantu

, and many others for tuning in. Subscribe to the ASP newsletter to make sure you don’t miss the next one!

Normalizing Political Retribution in America

With the news that the Trump administration is investigating James Comey and John Brennan, ostensibly for their role in investigating and informing the public about Russia’s 2016 election interference efforts, I wrote for MSNBC that “the pattern of politically motivated investigations that the Trump administration is pursuing is meant to desensitize us, to quell outrage and to stifle dissent.” I also wrote about the human consequences of targeting perceived “enemies”—consequences I know personally.

The same thing, of course, is happening on a much larger scale to the 1300+ State Department employees fired this week, the thousands of USAID employees officially terminated at the beginning of the month, and other targeted firings of civil servants who worked on the January 6 prosecutions, disinformation, election security, and beyond. Read the essay here.

Bad Actors are Using AI, Just Not the Way You Think

A few weeks ago I gave the opening keynote at the World Summit AI in San Francisco. My speech centered on the ways foreign adversaries are employing artificial intelligence to spread disinformation and influence audiences. While we’ve been told that we should expect world-altering deepfakes any day now, the reality is a lot more pernicious. You can watch my speech here.

The speech drew on ASP’s research on the Pravda network, which is grooming large language models to produce Russian propaganda. We tested the latest “reasoning” models of ChatGPT and found they’re still regurgitating this slop, almost six months after we released our landmark report. Read our latest, with my coauthors

Sophia Freuden

and AI visionary

Gary Marcus

:

Bad Actors are Grooming LLMs to Produce Falsehoods

Thanks to subscribers old and new for your support.