Marten Risius
@risius.bsky.social
Professor for Digital Society and Online Engagement, @dsrc.bsky.social
Trust and Safety at University of Applied Sciences Neu-Ulm, views are my own
Trust and Safety at University of Applied Sciences Neu-Ulm, views are my own
What is the gossip on why X is releasing new content moderation features? Seems off-brand to me. What am I missing?
www.techspot.com/news/110332-...
www.techspot.com/news/110332-...
X wants to call you out for using a VPN (and maybe catch a few trolls, too)
The situation began in October when Mikita Bier, X's Head of Product, announced that new information would be shown on user profiles, such as the date they...
www.techspot.com
November 25, 2025 at 12:32 PM
What is the gossip on why X is releasing new content moderation features? Seems off-brand to me. What am I missing?
www.techspot.com/news/110332-...
www.techspot.com/news/110332-...
@riekers.bsky.social is doing awesome work using multi-agent systems, in this case to detect hate speech in memes. Quite a challenging task but a promising approach and much more to come.
Reach out if you're interested or struggling to find explainable AI solutions for multimodal applications
Reach out if you're interested or struggling to find explainable AI solutions for multimodal applications
Do multiple perspectives improve hate speech detection?
Presented at #ACMMM25 by our colleague @riekers.bsky.social, our new framework "MAXplain" introduces a multi-agent approach to multimodal hate speech detection. 1/6
🔗 dl.acm.org/doi/10.1145/...
Presented at #ACMMM25 by our colleague @riekers.bsky.social, our new framework "MAXplain" introduces a multi-agent approach to multimodal hate speech detection. 1/6
🔗 dl.acm.org/doi/10.1145/...
MAXplain: A Multi-Agent System for Interactive Multimodal Hate Speech Detection | Proceedings of the 33rd ACM International Conference on Multimedia
dl.acm.org
November 4, 2025 at 1:59 PM
@riekers.bsky.social is doing awesome work using multi-agent systems, in this case to detect hate speech in memes. Quite a challenging task but a promising approach and much more to come.
Reach out if you're interested or struggling to find explainable AI solutions for multimodal applications
Reach out if you're interested or struggling to find explainable AI solutions for multimodal applications
Reposted by Marten Risius
We have a new preprint: osf.io/preprints/so...
What have we learned about social media - the constantly moving target of empirical research - over the past decade?
What have we learned about social media - the constantly moving target of empirical research - over the past decade?
October 30, 2025 at 10:53 AM
We have a new preprint: osf.io/preprints/so...
What have we learned about social media - the constantly moving target of empirical research - over the past decade?
What have we learned about social media - the constantly moving target of empirical research - over the past decade?
While prebunking has mostly been used in the disinformation context, @marcoduerr.bsky.social explores its potential for dealing with other kinds of harm - in this case authoritarian attitudes. While these preliminary results are promising, we work to further refine these interventions.
What if we could counter authoritarian and anti-democratic attitudes before they take root?
Our colleague @marcoduerr.bsky.social proposes attitude inoculation as a preventive strategy to build resistance against harmful political mindsets.
aisel.aisnet.org/icis2025/is_... 1/6
Our colleague @marcoduerr.bsky.social proposes attitude inoculation as a preventive strategy to build resistance against harmful political mindsets.
aisel.aisnet.org/icis2025/is_... 1/6
AIS Electronic Library (AISeL) - ICIS 2025 Proceedings: Mitigating the Effects of Harmful Social Media Content through Attitude Inoculation
Online social media platforms enable the swift dissemination of harmful content, such as hate speech, misinformation, and extremist propaganda, that threatens democratic values and fosters political c...
aisel.aisnet.org
October 23, 2025 at 10:05 AM
While prebunking has mostly been used in the disinformation context, @marcoduerr.bsky.social explores its potential for dealing with other kinds of harm - in this case authoritarian attitudes. While these preliminary results are promising, we work to further refine these interventions.
Soooooo ... are we now embracing research?... 👁️
October 20, 2025 at 7:51 PM
Soooooo ... are we now embracing research?... 👁️
Published 2025 but maybe already outdated? The rapid advance of AI and the fundamental shifts to our (increasingly hyperpersonalized) information environment, we might have to revisit this sooner than expected.
Are you in an echo chamber or stuck in your own information cocoon?
In a recent paper co-written by our @risius.bsky.social, they unpack the difference and why it matters for how we consume information.
Read the thread for details.
doi.org/10.1016/j.jsis.2025.101904
1/6
In a recent paper co-written by our @risius.bsky.social, they unpack the difference and why it matters for how we consume information.
Read the thread for details.
doi.org/10.1016/j.jsis.2025.101904
1/6
October 14, 2025 at 12:33 PM
Published 2025 but maybe already outdated? The rapid advance of AI and the fundamental shifts to our (increasingly hyperpersonalized) information environment, we might have to revisit this sooner than expected.
“The world will be Tlön.” - Jorge Luis Borges, Tlön, Uqbar, Orbis Tertius (1940)
Good job Sora 2 has watermarks on its videos (people have already made tools that instantly remove the watermark)
www.404media.co/sora-2-water...
www.404media.co/sora-2-water...
Sora 2 Watermark Removers Flood the Web
Bypassing Sora 2's rudimentary safety features is easy and experts worry it'll lead to a new era of scams and disinformation.
www.404media.co
October 8, 2025 at 1:33 PM
“The world will be Tlön.” - Jorge Luis Borges, Tlön, Uqbar, Orbis Tertius (1940)
This is where we will share most of our work, for anyone interested in #TrustandSafety reserach
Hello Bluesky! We are a research collaboratory at Neu-Ulm University of Applied Sciences in Germany. 1/3
October 8, 2025 at 1:05 PM
This is where we will share most of our work, for anyone interested in #TrustandSafety reserach
It's a correlation, but it might indicate the value of proper content moderation
Have we passed peak social media? @jburnmurdoch.ft.com here www.ft.com/content/a072...
October 3, 2025 at 1:39 PM
It's a correlation, but it might indicate the value of proper content moderation
No welcome from Tom Anderson?... I don't know about this
a picture of a smiling man with the name tom on top
ALT: a picture of a smiling man with the name tom on top
media.tenor.com
September 29, 2025 at 6:07 AM
No welcome from Tom Anderson?... I don't know about this