In May 2023, the President’s Council of Advisors on Science and Technology (PCAST) launched a working group on generative artificial intelligence (AI) and invited public input. Integrity Institute visiting fellows Theodora Skeadas, David Evan Harris, and Arushi Saxena organized and – together with Institute members Diane Chang and Sabhanaz Rashid Diya – submitted comments to the PCAST.

In April 2023, select Institute members addressed three main topics in their comments on the draft EU AI Act: examples of real life harms that have resulted from the use of AI systems; the categories of AI systems classified as “high risk”; and a draft methodology for auditing AI systems.

Select Institute members share their thoughts on the potential of generative AI in integrity, trust & safety work

Select Institute members share their thoughts on why AI may make integrity jobs harder

Integrity Institute founding fellow Tom Cunningham shares his six observations on ranking by engagement on social platforms and – via original research – visualizes how platforms are trading-off between different outcomes.

In March 2023, the Oversight Board accepted Meta’s request for a policy advisory opinion on the company’s approach to moderating the Arabic term “shaheed” and solicited public comments. Select Institute members prepared these comments and the Institute submitted them to the Oversight Board on their behalf.

Integrity Institute member Theodora Skeadas offers an alternative list to the New York Times opinion column “How Twitter Changed the World, in 25 Tweets," putting global events and narratives at the center.

2024 has 83 elections globally across 78 countries. We won’t see that many again until 2048. Integrity Institute fellow Katie Harbath and co-author Ana Khizanishvili share what these numbers from global elections tell us about the future of democracy.

Integrity Institute fellow Dylan Moses argues the importance of Section 230 and explains the legal nuts and bolts of Gonzalez and Taamneh.

Integrity Institute fellow Karan Lala argues that the nature of the internet makes moderating every bit of speech impossible. Holding platforms responsible for every dangerous statement that slips through their nets will have a ruinous chilling effect on modern discourse.

We have analyzed the misinformation amplification factor for Facebook and Twitter ahead of the Nigerian election. We find the trend we saw in the US, with both platforms significantly amplifying misinformation and Twitter more so than Facebook.

Reporting from the WSJ on internal discussions at Facebook show internally, Facebook considers the top content lists this quarter to have “no low quality content.” Contrary to that, we see little change in the number of posts that fail basic media literacy checks.

Integrity Institute members Alice Hunsberger, Kristen Murdock, and Jen Weedon offer their guide to preventing & reacting to burnout – an issue that impacts many working professionals but may hit those working in integrity, trust & safety especially hard.

For the 2022 US midterm elections, we have been tracking how misinformation is amplified on large platforms to show how much additional engagement one should expect to get on the platforms for posting misinformation.

We are tracking how misinformation performs on platforms to measure the extent to which platforms are amplifying misinformation, and the extent to which they are creating an incentive structure that rewards lies and misinformation online.

Explaining middleware, a new approach to content moderation that puts power back in the hands of users

The Integrity Institutes analysis of Facebook's Widely Viewed Content Report for Q2 2022. We dig into the report to extract every possible insight and piece of data into how Facebook is performing and to what extent harmful or spammy content is making it to their top lists.

Social Spam is backed by a microcosm of the modern economy. In order to understand this world, I infiltrated one of the largest communities for spammers with my real Facebook account. Here’s what I learned.

When studying content and behavior violations on platforms, different surfaces with equal numbers of consumers are not necessarily alike