🚨 New preprint 🚨 Can AI agents coordinate influence campaigns without human guidance? And how does coordination arise among AI agents? In our latest research, we simulate LLM-powered AI agents acting like users on an online platform, some benign, some running an influence operation

New paper + interactive dashboard on the 2024 election information ecosystem. Building on discourse networks work with @hanshanley.bsky.social @luceriluc.bsky.social @emilioferrara.bsky.social—this lets us visualize the online landscape as a unified system, rather than isolating each platform.

Thrilled to share our latest paper "Information Suppression in Large Language Models" is now published on Information Sciences! To read more, see: www.sciencedirect.com/science/arti... great work w/ @siyizhou.bsky.social

Peiran Qiu, Siyi Zhou, Emilio Ferrara: Information Suppression in Large Language Models: Auditing, Quantifying, and Characterizing Censorship in DeepSeek https://arxiv.org/abs/2506.12349 https://arxiv.org/pdf/2506.12349 https://arxiv.org/html/2506.12349

Amin Banayeeanzade, Ala N. Tak, Fatemeh Bahrani, Anahita Bolourani, Leonardo Blas, Emilio Ferrara, Jonathan Gratch, Sai Praneeth Karimireddy Psychological Steering in LLMs: An Evaluation of Effectiveness and Trustworthiness https://arxiv.org/abs/2510.04484

Big news from #ICWSM2025! "The Susceptibility Paradox in Online Social Influence” by @luceriluc.bsky.social, @jinyiye.bsky.social, Julie Jiang & @emilioferrara.bsky.social was named a Top 5 Paper & won Best Paper Honorable Mention! 👏 Congrats to all!

So satisfying to have some evidence that Elon Musk is wildly promoting himself on X. Researchers made 120 sock puppet accounts to see whose content is getting pushed on users. #FAccT2025 @jinyiye.bsky.social @luceriluc.bsky.social @emilioferrara.bsky.social doi.org/10.1145/3715...

A Multimodal TikTok Dataset of #Ecuador's 2024 Political Crisis and Organized Crime Discourse by Gabriela Pinto, @emilioferrara.bsky.social USC workshop-proceedings.icwsm.org/abstract.php... @icwsm.bsky.social #dataforvulnerable25

🤖Thrilled to share our latest work☄️ Have you ever wondered what LLMs know but they are not saying? We built an auditing framework to study information suppression in LLMs, and demonstrated it to quantify and characterize censorship in DeepSeek. Read more: arxiv.org/abs/2506.12349

I'll be at #ICWSM 2025 next week to present our paper about Bluesky Starter Packs. For the occasion, I've created a Starter Pack with all the organizers, speakers, and authors of this year I could find on Bluesky! Link: go.bsky.app/GDkQ3y7 Let me know if I missed anyone!

Paper here: arxiv.org/abs/2505.21729

🚨 𝐖𝐡𝐚𝐭 𝐡𝐚𝐩𝐩𝐞𝐧𝐬 𝐰𝐡𝐞𝐧 𝐭𝐡𝐞 𝐜𝐫𝐨𝐰𝐝 𝐛𝐞𝐜𝐨𝐦𝐞𝐬 𝐭𝐡𝐞 𝐟𝐚𝐜𝐭-𝐜𝐡𝐞𝐜𝐤𝐞𝐫? new "Community Moderation and the New Epistemology of Fact Checking on Social Media" with I Augenstein, M Bakker, T. Chakraborty, D. Corney, E Ferrara, I Gurevych, S Hale, E Hovy, H Ji, I Larraz, F Menczer, P Nakov, D Sahnan, G Warren, G Zagni

One of my favorite recent projects! Link to the paper: arxiv.org/abs/2505.10867

What does coordinated inauthentic behavior look like on TikTok? We introduce a new framework for detecting coordination in video-first platforms, uncovering influence campaigns using synthetic voices, split-screen tactics, and cross-account duplication. 📄https://arxiv.org/abs/2505.10867

"Limited effectiveness of LLM-based data augmentation for COVID-19 misinformation stance detection" by @euncheolchoi.bsky.social @emilioferrara.bsky.social et al, presented by the awesome Chur at The Web Conference 2025 arxiv.org/abs/2503.02328

🚀 First study on multimodal AI-generated content (AIGC) on social media! TLDR: AI-generated images are 10× more prevalent than AI-generated text! Just 3% of text spreaders and 10% of image spreaders drive 80% of AIGC diffusion, with premium & bot accounts playing a key role🤖📢

🤩Cool collaboration w/ @jinyiye.bsky.social @emilioferrara.bsky.social @luceriluc.bsky.social 🔍Read more: arxiv.org/abs/2502.11248 📊Resources available: github.com/angelayejiny...

How does DeepSeek censorship work? Here is a practical example: I asked it to discuss about my work (having studied censorship online by various countries). DeepSeek at first starts to compose an accurate answer, even mentioning China’s online censorship efforts.

Good day to boost this paper on the contrasting effects of lax content moderation (like on X and what is coming at Meta) and how they drive toxic content; by a team including @luceriluc.bsky.social and @emilioferrara.bsky.social arxiv.org/abs/2412.157...

i was annoyed at having many chrome tabs with PDF papers having uninformative titles, so i created a small chrome extension to fix it. i'm using it for a while now, works well. today i put it on github. enjoy. github.com/yoavg/pdf-ta...

When I was a grad student I looked up to giants of my discipline and never would have thought a day like this would ever come for me, happy to celebrate a personal milestone on this holiday! Academia and research are great. Thanks all!

"IOHunter: Graph Foundation Model to Uncover Online Information Operations" goes to AAAI'25! This is the result of an incredible collaboration with @luceriluc.bsky.social @frafabbri.bsky.social and @emilioferrara.bsky.social Read the entire thread for a summary and the link to the preprint.

Fascinating new look at source selection in news from Spangher, @emilioferrara.bsky.social et al. doi.org/10.48550/arX...

An incredibly important work by Jinyi, Luca and Julie! #icwsm2025 Thread below:

New evidence of cross-platform foreign interference on social media during the 2024 U.S. Election that drove the spread of highly-partisan, low-credibility, and conspiratorial content, from Cinus, Minici, @luceriluc.bsky.social @emilioferrara.bsky.social arxiv.org/pdf/2410.22716

That's an amazing recognition, proud of Luca!

A rent paper has been portrayed in the news as proof that Twitter/X new #recsys favors republicans. In our latest work, collecting millions of datapoints via 120 sock puppets, we should the reality is much more complex: you would be surprised by our findings: arxiv.org/abs/2411.01852 thread 1/n

Ready for another Computational Social Science Starter Pack? Here is number 2! More amazing folks to follow! Many students and the next gen represented! go.bsky.app/GoEyD7d

What are the risks of living in a GenAI synthetic reality? I’ve been iterating over some (still fuzzy) ideas about one of my deepest fears for what the future holds. Still not polished enough for a submission but I’m happy to get feedback! arxiv.org/abs/2411.08250 #genai #css

Sharing my first Computational Social Science starter pack! Will grow with time, feel free to nominate and self nominate! go.bsky.app/CYmRvcK

#Data2thePeople! Pls reshare! Do you want a billion telegram posts?! You get data! You get data! You get data! Paper arxiv.org/abs/2410.23638 Data github.com/leonardo-bla... #election2024

🔥 More #data2thepeople! #election2024 🔥 You'll need a fire hydrant to fight all the data that's coming out our lab! How about the entire Truth Social election-related chatter? Anybody wants it? There you have it! Paper arxiv.org/abs/2411.01330 Dataset github.com/kashish-s/Tr... Pls reshare!

#election2024 Dataset update! #data2thePeople We have just released a new version of our TikTok data that goes up to November 1! 3+ million election related TikTok’s! Paper arxiv.org/abs/2407.01471 Data github.com/gabbypinto/U... Pls repost with the computational and social science communities!

More #data2thepeople! #election2024 We just released a Twitter/X dataset of 22m+ tweets about the upcoming election! (Data will be routinely updated, keep an eye and spread the word!) Paper arxiv.org/abs/2411.00376 Data github.com/sinking8/usc...

We are seeing new formats, new modalities of manipulation of some sort including generative AI, the use of these mock news websites to preach more fringe stories and, most importantly perhaps, the fact that now these campaigns span the entire media ecosystem online thehill.com/policy/techn...

“Once it’s being incubated in those fringe environments and it becomes relevant enough, the bridge . . . cross-pollinates the mainstream platform, like X, with the content,” Ferrara says. www.fastcompany.com/91221700/usc...

We discovered a network of coordinated inauthentic actors using AI-content and cross-platform link-sharing to manipulate the online discourse of the 2024 U.S. Presidential election. The network spans X and YouTube, disseminating biased political content through fake news. arxiv.org/abs/2409.15402

Some folks did a survey of surveys on fairness :) arxiv.org/abs/2403.17333 It turns out that my survey is ranked as the most complete, so perhaps you wanna take a look at it too! mdpi.com/2413-4155/6/1/3

Finishing the year on a high note!

Coverage of our latest work! Online ‘likes’ for toxic social media posts prompt more − and more hateful − messages https://theconversation.com/online-likes-for-toxic-social-media-posts-prompt-more-and-more-hateful-messages-218220

MASSIVE day for our lab! Julie Jiang and Herbert Chang made the Forbes 30 Under 30 list in the Science category! I couldn't be prouder, so richly deserved! www.forbes.com/30-under-30/...

Just joined yup! Finally Bluesky, Threads, and Twitter in one app.

📝🫳⚠️ Paper Drop Alert 🔥 Quick, tell all your friends: Should ChatGPT be biased? Challenges and risks of bias in large language models Finally out on First Monday! firstmonday.org/ojs/index.ph...

Come join us at the Thomas Lord Department of Computer Science at USC! Searchers for tenure track faculty in HCI, AI/ML/Data Science, and privacy/security! usccareers.usc.edu/job/los-ange...

Some of our lab’s work linked in the story!

Why is there so much toxicity and hate online? It may be because people are seeking social approval from their social networks (not that they directly want to harm the target). Cool paper by Jiang, Luceri, Walther & @emilioferrara.bsky.social doi.org/10.48550/arX...

❤️‍🔥Here is a new paper I am really excited about🔥 Social Approval and Network Homophily as Motivators of Online Toxicity arxiv.org/abs/2310.07779 Online hate is driven by the pursuit of social approval: toxicity is homophilous and a user's propensity for it can be predicted by their social networks!

Our latest work is out on EPJ Data Science! Exposing influence campaigns in the age of LLMs: a behavioral-based AI approach to detecting state-sponsored trolls epjdatascience.springeropen.com/articles/10....

We studied this very question in a 2017 paper: "The misleading narrative of the canonical faculty productivity trajectory"! The canonical trajectory (what @carlbergstrom.com screen capped) holds for only about 25-30% of faculty -- variety is the norm! www.pnas.org/doi/abs/10.1...