Sign in

Mohsen Mosleh

@mmosleh.bsky.social
3.5K followers 674 following 22 posts

Prof @Oxford, Affiliate @MIT www.MohsenMosleh.com

PostsRepliesMedia
Mohsen Mosleh @mmosleh.bsky.social · 23/09/2026
8/ location transparency works through the disclosed accounts, not their audiences. Preprint: arxiv.org/abs/2609.25933
arxiv.org
Location transparency reduces activity by accounts misrepresenting their location on X
Concerns about inauthentic accounts, including foreign actors posing as domestic voices, are central to debates about online discourse. Yet, little is known about accounts with inaccurate location cla...
080 8
Mohsen Mosleh @mmosleh.bsky.social · 23/09/2026
7/ And the audience? Nothing. Reposts and replies per post were unchanged. X shows location on the profile page, not beside posts, so most readers never see it. Total engagement fell 14% — but only because the accounts posted less, not because readers turned away.
170 7
Mohsen Mosleh @mmosleh.bsky.social · 23/09/2026
6/ Disclosure also changed whom they talk to. Replies from mismatched accounts to US-based users fell 10%; toxic replies to US users fell 18%. Replies to non-US users: no change. The restraint targets exactly the audience whose location they had claimed to share.
170 7
Mohsen Mosleh @mmosleh.bsky.social · 23/09/2026
5/ Who pulls back most? VPN-flagged accounts (−26%), accounts with prior username changes (−22%), non-verified accounts (−21%), and accounts heavy on crypto and scam content. That looks more like operations losing their cover than general discomfort at being outed.
170 7
Mohsen Mosleh @mmosleh.bsky.social · 23/09/2026
4/ The effect is far from uniform. Accounts revealed to be in Africa dropped 29.2%, Asia 24.4%. Europe, non-US North America, South America and Oceania: small or null. Where genuine relocation is plausible, little happens; where it isn't, activity falls sharply.
170 7
Mohsen Mosleh @mmosleh.bsky.social · 23/09/2026
3/ Difference-in-differences over 21 weeks (1.3M posts, 3.6M replies): disclosure cut mismatched accounts' posting activity by 13.1%. Pre-trends are flat, and the effect holds under synthetic DiD, propensity score matching, daily-level estimates and X's own API counts.
160 6
Mohsen Mosleh @mmosleh.bsky.social · 23/09/2026
2/ Before disclosure, mismatched accounts already looked different: newer, fewer followers, more username changes and VPN flags, and more scam, crypto and misleading content. But no partisan tilt — their left- and right-leaning shares matched those of US-based accounts.
170 7
Mohsen Mosleh @mmosleh.bsky.social · 23/09/2026
New paper: In Nov 2025, X began showing where accounts actually operate from. We studied 8,200 politically engaged accounts claiming to be in the US half were revealed to be elsewhere What happened when the mismatch went public? @yuweichuai.bsky.social @thomasrenault.bsky.social @dgrand.bsky.social
16018 60
Mohsen Mosleh @mmosleh.bsky.social · 31/08/2026
CN can constrain misleading posts w/o durably changing the accounts that produce most of them. Correcting content and changing its producers are distinct platform-design objectives. Work led by @yuweichuai.bsky.social together w/ @thomasrenault.bsky.social , nicp.bsky.social, Gabriele Lenzini 🙏
010 1
Mohsen Mosleh @mmosleh.bsky.social · 31/08/2026
Design implications: Post-level metrics overstate effectiveness — author-level outcomes matter too The first correction is the key intervention point; later notes barely register Showing the same note format repeatedly doesn't move persistent offenders
100 1
Mohsen Mosleh @mmosleh.bsky.social · 31/08/2026
And after correction, the content divergence is stark: Once-corrected accounts → less toxic, less misleading posts ✅ Repeatedly corrected accounts → lower-quality sources, more political content, and more highly misleading posts overall ❌
111 1
Mohsen Mosleh @mmosleh.bsky.social · 31/08/2026
Who are they? Before any correction, repeatedly corrected accounts already have more followers, are more likely to be verified, lean more right, and post more frequent, more misleading, more political content.
100 1
Mohsen Mosleh @mmosleh.bsky.social · 31/08/2026
Why does this matter? Concentration. Repeatedly corrected accounts are only 28.6% of corrected accounts — but they produce 73.4% of all fact-checked posts. The accounts most responsible for flagged misinformation are the ones least responsive to correction.
100 1
Mohsen Mosleh @mmosleh.bsky.social · 31/08/2026
Accounts corrected repeatedly? They post ↑4.4% after their first correction — and show no detectable response to any correction after that. The more notes an account eventually receives, the bigger that initial jump (up to +9.5% for accounts with 5+ notes).
100 1
Mohsen Mosleh @mmosleh.bsky.social · 31/08/2026
But that average hides two completely different stories. Accounts corrected only once (71% of corrected accounts): post ↓2.4%, and their subsequent content is less toxic and less misleading. Corrective learning seems to work.
100 1
Mohsen Mosleh @mmosleh.bsky.social · 31/08/2026
First, the surprise: on average, accounts post 2.9% MORE after being corrected — an effect that emerges immediately and persists for the full four weeks.
100 1
Mohsen Mosleh @mmosleh.bsky.social · 31/08/2026
New preprint: Community Notes are good at correcting posts. But do they change the people behind them? We tracked 19,854 corrected accounts on X — 11.9M posts, 4 weeks before/after each correction. The answer: it depends who you are. @oii.ox.ac.uk 📎 arxiv.org/abs/2608.27526
arxiv.org
Community corrections have divergent downstream effects across corrected accounts
Community-based fact-checking can reduce the spread of annotated misleading posts, but whether it produces lasting behavioral change among corrected authors remains unclear. Here, we conduct a large-s...
2158 15
Reposted by Mohsen Mosleh
Basak Bozkurt @basakbozkurt.bsky.social · 22/07/2026
1/ New in HKS Misinformation Review: as platforms turn to Community Notes, what do fact-checkers themselves think? We interviewed 29 fact-checkers across 41 countries. w/ @mmosleh.bsky.social & @helenmargetts.bsky.social 📄 misinforeview.hks.harvard.edu/article/fact...
misinforeview.hks.harvard.edu
Fact-checking at a crossroads: Fact checkers’ perspectives on Community Notes, AI integration, and design recommendations | HKS Misinformation Review
Social media platforms are increasingly using community-based verification systems, such as Community Notes, and AI systems to flag and contextualize potentially misleading content at scale. While the...
111 1
Reposted by Mohsen Mosleh
Basak Bozkurt @basakbozkurt.bsky.social · 12/06/2026
Excited to be organising a workshop at #GlobalFact2026 in Vilnius with @mmosleh.bsky.social, Naomi Miyashita, and Joseph Aylett-Bullock supported by the @oxpolicyengaged.bsky.social Seed Fund.
112 2
Reposted by Mohsen Mosleh
David Rand @dgrand.bsky.social · 04/02/2026
Grok fact-checks our paper on Grok fact-checking - and it approves!
1297 29
Reposted by Mohsen Mosleh
David Rand @dgrand.bsky.social · 03/02/2026
🚨New WP "@Grok is this true?" We analyze 1.6M factcheck requests on X (grok & Perplexity) 📌Usage is polarized, Grok users more likely to be Reps 📌BUT Rep posts rated as false more often—even by Grok 📌Bot agreement with factchecks is OK but not great; APIs match fact-checkers osf.io/preprints/ps...
211747 117
Reposted by Mohsen Mosleh
Alexios Mantzarlis @mantzarlis.com · 28/01/2026
In a new working paper, @thomasrenault.bsky.social, @mmosleh.bsky.social & @dgrand.bsky.social break down how Grok is being used to fact-check, and on whom. osf.io/preprints/ps...
172 7
Reposted by Mohsen Mosleh
Alexios Mantzarlis @mantzarlis.com · 28/01/2026
New on @indicator.media: "@grok is this true" was the single most frequent reply tagging X's AI chatbot in the six months following its launch.
indicator.media
@Grok is this true: How X’s chatbot performs as a fact-checking tool
New research explores whether the chatbot might replace the crowdsourced fact-checking program – and what that might mean for getting to the truth on X
22910 29
Reposted by Mohsen Mosleh
Oxford Internet Institute @oii.ox.ac.uk · 21/11/2025
The workshop, organised by the OII's Dr @mmosleh.bsky.social, Dylan Thurgood and Manuel Tonneau, featured keynotes, lightning/research talks, and a policy session. Huge thanks to our organisers, the attendees, and our fantastic session chairs for making the event such a success!
021 2
Reposted by Mohsen Mosleh
Oxford Internet Institute @oii.ox.ac.uk · 21/11/2025
50 researchers from across Europe and North America recently came to University of Oxford for a two-day workshop at St Anne's College, University of Oxford, exploring the future of social media research.
141 4
Reposted by Mohsen Mosleh
Dr Moira Nicolson @moiranics.bsky.social · 17/11/2025
Amazing new paper analysed 10 million news posts on 7 social media platforms in January 2024 - they found that most content posted is high quality and politically moderate, including on Twitter/X. The problem is that engagement is much higher for the relatively small portion of low quality content.
Frequency distribution of political lean of news domains shared on BlueSkye, Gab, Gettr, LinkedIn, Mastodon, TruthSocial, x/twitter. The most common type of content on Bluesky, LinkedIn, Twitter and Mastodon is politically moderate. For TruthSocial, the most common content is right wing. Gettr and Gab is varied. Frequency distribution of quality of news shared on each platform. Most content on Bluesky, X/Twitter, Mastodon and LinkedIn is of medium to high quality. Gab, Gettr and Truth Social is of varying quality. Line graphs showing the proportion of posts that are among the 10% highest engagement posts on each platform by news quality (binned by rounding quality to nearest 0.25). Shows that for all platforms, except BlueSky, engagement declines as quality increases, especially for LinkedIn. BlueSky sees the highest engagement for posts of average quality.
196 9
Reposted by Mohsen Mosleh
David Rand @dgrand.bsky.social · 14/11/2025
🚨Out in PNAS🚨 Examining news on 7 platforms: 1)Right-leaning platforms=lower quality news 2)Echo-platforms: Right-leaning news gets more engagement on right-leaning platforms, vice-versa for left-leaning 3)Low-quality news gets more engagement EVERYWHERE - even BlueSky! www.pnas.org/doi/10.1073/...
11216104 216
Reposted by Mohsen Mosleh
Mark Robinson @mr.place · 13/11/2025
A joy of later life studies @lsepbs.bsky.social is the dept. lectures @mmosleh.bsky.social on news quality & political bias on social media was excellent. www.pnas.org/doi/epdf/10.... - x is slightly liberal leaning😱 Crap news travels fast on the right. @t0nyyates.bsky.social have you seen this?
pnas.org
PNAS
Proceedings of the National Academy of Sciences (PNAS), a peer reviewed journal of the National Academy of Sciences (NAS) - an authoritative source of high-impact, original research that broadly spans...
011 1
Reposted by Mohsen Mosleh
Brendan Nyhan @brendannyhan.bsky.social · 05/11/2025
"While high-quality content is posted more and receives more total engagement across platforms...a given author attracts higher levels of engagement when they post lower-quality content" "pattern we find seems to be driven more by an underperformance of particularly popular high-quality outlets"
pnas.org
Divergent patterns of engagement with partisan and low-quality news across seven social media platforms | PNAS
In recent years, social media has become increasingly fragmented, as platforms evolve and new alternatives emerge. Yet most research studies a sing...
23913 39
Reposted by Mohsen Mosleh
Jenny Allen @jennyallen.bsky.social · 05/11/2025
Now in PNAS! With @mmosleh.bsky.social and @dgrand.bsky.social Examining news engagement across 7 platforms, we find: - differences WRT political lean ("echo platforms") - similarities WRT news quality (posting low qual news => more engagement) Read paper and thread for more!
pnas.org
PNAS
Proceedings of the National Academy of Sciences (PNAS), a peer reviewed journal of the National Academy of Sciences (NAS) - an authoritative source of high-impact, original research that broadly spans...
1243 24
Reposted by Mohsen Mosleh
Oxford Internet Institute @oii.ox.ac.uk · 03/11/2025
New! Great piece by Tech and Science Post highlighting new research study examining partisan news sharing on social media platforms, lead author Dr @mmosleh.bsky.social, @oii.ox.ac.uk, co-authors Jennifer Allen and @dgrand.bsky.social. 1/2 techandsciencepost.com/news/other-s...
techandsciencepost.com
Low-quality news links draw higher engagement, no matter the political slant online - Tech and Science Post
The spread of dubious headlines on social media isn’t just a right-wing thing—it’s a social media thing, according to new research from Cornell. After studying millions of social media posts containin...
1104 10
Reposted by Mohsen Mosleh
Oxford Internet Institute @oii.ox.ac.uk · 31/10/2025
NEW: In a recent Oxford-led study by @mmosleh.bsky.social, researchers analysed millions of social media posts containing links to news stories, across seven different social platforms. Read the full research paper here: www.pnas.org/doi/10.1073/...
277 7
Reposted by Mohsen Mosleh
Oxford Internet Institute @oii.ox.ac.uk · 02/10/2025
According to @mmosleh.bsky.social and colleagues, information integrity can learn from the threat intelligence practices routinely utilized in cybersecurity. Read the full @weforum.org piece here: www.weforum.org/stories/2025...
weforum.org
Why cybersecurity and information integrity are two sides of the same coin
With cybercriminals increasingly employing both cybersecurity and information-integrity-related threats, both disciplines are now integrally linked.
021 2
Reposted by Mohsen Mosleh
Sarah Shugars @shugars.bsky.social · 13/08/2025
So many interesting findings in this work from @mmosleh.bsky.social and @dgrand.bsky.social comparing news sharing and engagement by quality and partisanship across 7 platforms Notable partisan asymmetries. Low quality news gets more engagement, driven by traditional news getting poor engagement
021 2
Reposted by Mohsen Mosleh
Meysam Alizadeh @malizadeh.bsky.social · 18/07/2025
🚨 New preprint alert: We show that web-browsing GPT and LLaMA models can infer social media user demographics with reasonable accuracy—using only usernames. This opens new possibilities for social media research in the post-API era but raises important privacy concerns.
1115 11
Reposted by Mohsen Mosleh
Oxford Internet Institute @oii.ox.ac.uk · 16/07/2025
New research! Experts from @oii.ox.ac.uk and @mitsloan.bsky.social explore how friends of friends become friends on social media platform X in new research published in @pnas.org. 1/4
143 4
Reposted by Mohsen Mosleh
Wolfson College Oxford @wolfsonoxford.bsky.social · 02/07/2025
Wolfson GBF Prof @mmosleh.bsky.social has co-authored a study revealing that posts made by Republican users on X are more than twice as likely to be flagged for misinformation than those by Democrats. Read more at our website: www.wolfson.ox.ac.uk/news/misinfo...
wolfson.ox.ac.uk
Republican Tweets More Likely to be Flagged for Misinformation
New study by Wolfson Fellow finds Republicans’ posts are flagged twice as often as Democrats for misinformation on X’s Community Notes.
021 2
Reposted by Mohsen Mosleh
David Rand @dgrand.bsky.social · 17/06/2025
🚨In PNAS🚨 The right often accuses fact-checkers of political bias But we analyzed Community Notes on Musk's X and found posts flagged as "misleading" are 2.3x more likely to be written by Reps than Dems! The issue is Reps sharing misinformation, not fact-checker bias... www.pnas.org/doi/10.1073/...
7344129 344
Reposted by Mohsen Mosleh
Thomas Renault @thomasrenault.bsky.social · 17/06/2025
🚨 New in PNAS 🚨 Posts by Republicans are 2.3 times more likely to be flagged as misleading than those by Democrats on X's Community Notes. A 🧵 pnas.org/doi/epub/10.... (Open Access)
176 7
Mohsen Mosleh @mmosleh.bsky.social · 14/05/2025
Happy to help. Just send me an email!
110 1
Reposted by Mohsen Mosleh
Oxford Internet Institute @oii.ox.ac.uk · 24/04/2025
In case you missed it, OII Prof @mmosleh.bsky.social has authored an expert comment piece for @ox.ac.uk on his recent research on how to increase engagement with factchecking. Read it here: www.ox.ac.uk/news/2025-04...
ox.ac.uk
Expert Comment: How can we encourage engagement with online
Dr Mohsen Mosleh, Associate Professor in Social Data Science at Oxford Internet Institute, explores the role of shared ideology and social connection in whether people engage with fact-checks that
032 3
Reposted by Mohsen Mosleh
Oxford Internet Institute @oii.ox.ac.uk · 22/04/2025
New! In his latest opinion piece, Associate Professor, @mmosleh.bsky.social @oii.ox.ac.uk asks how we can encourage engagement with online fact-checking? Read the full article: www.oii.ox.ac.uk/news-events/...
oii.ox.ac.uk
OII | Expert Comment: How can we encourage engagement with online fact-checking
In his latest opinion piece, Dr Mohsen Mosleh explores the role of shared ideology and social connection as influencing factors in fact-checking and countering online misinformation.
042 4
Reposted by Mohsen Mosleh
Wolfson College Oxford @wolfsonoxford.bsky.social · 22/04/2025
An exploration of the increasingly muddy waters of online fact-checking - a fascinating read from Wolfson GBF @mmosleh.bsky.social as he examines the role of shared ideology and social connection in whether people engage with fact-checks that aim to counter online misinformation.
052 5
Reposted by Mohsen Mosleh
Cameron Martel @cameronmartel.bsky.social · 14/04/2025
🚨New in @plosone.org🚨 Corrections of misinfo are often ignored. What can drive engagement? Twitter field exp & survey followups find -Social ties matter: users more likely to engage w corrections from accounts who followed user -Shared partisanship had smaller effects on engagement shorturl.at/0Ycdp
11912 19
Reposted by Mohsen Mosleh
Oxford Internet Institute @oii.ox.ac.uk · 01/04/2025
Read the full study ‘Promoting engagement with social fact-checks online: investigating the roles of social connection and shared partnership’, authors @cameronmartel.bsky.social . @mmosleh.bsky.social , @eckles.bsky.social and @dgrand.bsky.social. Download here: bit.ly/4iR4HbO
bit.ly
Promoting engagement with social fact-checks online: Investigating the roles of social connection and shared partisanship
Social corrections – where users correct each other – can help rectify inaccurate beliefs. However, social corrections are often ignored. Here we ask under what conditions social corrections promote e...
073 7
Reposted by Mohsen Mosleh
Oxford Internet Institute @oii.ox.ac.uk · 13/03/2025
In the latest episode of the OII podcast, we're tackling misinformation and polarization on social media. What's the real-world impact? How are governments responding? And what about AI and deepfakes? Listen here: podcasts.ox.ac.uk/why-social-m...
podcasts.ox.ac.uk
Why social media is the new frontier for misinformation, and what we can do about it: Professor Mohsen Mosleh and Cameron Martel
In the sixth episode of the OII Podcast, our experts discuss topics such as: * The real world impacts that arise when people increasingly identify with their political tribes online * What role govern...
185 8
Reposted by Mohsen Mosleh
Kobi Hackenburg @kobihackenburg.bsky.social · 07/03/2025
📈Out today in @PNASNews!📈 In a large pre-registered experiment (n=25,982), we find evidence that scaling the size of LLMs yields sharply diminishing persuasive returns for static political messages.  🧵:
14020 40
Reposted by Mohsen Mosleh
David Rand @dgrand.bsky.social · 31/01/2025
🚨New WP🚨 Remember Musk+Zuck+Trump+Jordan etc crying fact-checker bias b/c Reps were flagged more than Dems? We analyzed Community Notes on Musk's X and guess what: posts flagged as "misleading" are 67% more likely to be written by Reps! The issue is Reps, not fact-checkers... osf.io/preprints/ps...
6453121 453
Reposted by Mohsen Mosleh
Oxford Internet Institute @oii.ox.ac.uk · 27/01/2025
New paper alert! ‘Racial Minorities Face Discrimination From Across the Political Spectrum When Seeking to Form Ties on Social Media: Evidence From a Field Experiment’, authors: Krishnan Nair, @mmosleh.bsky.social @oiioxford.bsky.social; and Maryam Kouchaki. Download the study: bit.ly/3EbQnuZ
bit.ly
Sage Journals: Discover world-class research
Subscription and open access journals from Sage, the world's leading independent academic publisher.
052 5
Reposted by Mohsen Mosleh
Brendan Nyhan @brendannyhan.bsky.social · 07/01/2025
Fact-checkers have of course made mistakes, but perceptions of systematic bias are driven by a real-world asymmetry in misinformation sharing in the contemporary US that even politically balanced groups of laypeople recognize www.nature.com/articles/s41...
424664 246