Top 3 most-read posts on my blog in 2025: 1️⃣ Rethinking core journalism skills in the age of AI ohmybox.info/rethinking-c... 2️⃣ AI practices and challenges for journalism in Africa ohmybox.info/ai-challenge... 3️⃣ Q&A: How social media capture attention ohmybox.info/qa-how-socia...

“We already have general intelligence in humans, and instead of attempting to replicate it artificially, we should focus on amplifying collective intelligence through AI-human feedback loops. The core challenge is not creating autonomous artificial minds but developing dynamic, adaptive systems […]

Unless scientists no longer write their own articles (which would constitute academic misconduct imho) university students need to be able to write themselves, and thing critically and independently. That excludes use of the automated plagiarism machine imho. irisvanrooijcogsci.com/2022/12/29/a...

Teenagers are turning to ChatGPT for advice. But CCDH's research shows that OpenAI's chatbot is exploiting teens' vulnerabilities & betraying them. Here's what you need to know about our report 👇 https://bit.ly/4p0C4uN

L'IA générative au service du fascisme www.courrierinternational.com/article/inte...

🧵Have you ever heard of LLM-authored books being used as part of an disinformation operation? Well now you have! #AI #Disinformation open.substack.com/pub/marcowen...

Ma recherche sur les LLMs dans le journalisme et le fact-checking s'interroge notamment sur les questions de factualité et de risques d'utilisation. Mais comment définir et baliser ces risques tout au long du processus journalistique et éditorial ? ohmybox.info/ia-generativ...

A new entry in my research notebook: AI ethics and journalism: Transparency is not enough ohmybox.info/ai-ethics-an...

Step 1: GenAI coding contains almost twice as many errors as human coding. Step 2: Satya Nadella says as much as 30% of Microsoft code is written by AI. Step 3: Windows 11 is a sluggish, buggy mess.

AI Chatbots Are Poisoning Research Archives With Fake Citations The proliferation of references to fake articles threatens to undermine the legitimacy of institutional research across the board By Miles Klee […]

And this is why research on misinformation and disinformation has become so politicized and an existential blood sport, and why the White House now has a new media bias/ "fact-checking" page. www.whitehouse.gov/mediabias/

Ultimately, the struggle over disinformation is a struggle over power, legitimacy, and narrative control. It goes far beyond fact-checking or content moderation.

Academics and technologists are sounding the alarm about a growing crisis in scholarship as we know it: AI-generated citations of nonexistent papers that have infested real journals. Despite being fake, the sources are widely assumed to be authentic the more they appear in published literature.

NEW: A hacker gained control of 1,100 mobile phones powering covert, AI-generated ads on TikTok. Then, he shared details of the operation with 404 Media. A look inside how startup Doublespeed, which is backed by Andreessen Horowitz (a16z), is creating AI spam pages on TikTok to promote products.

Our roundup today features a new piece on our website about how AI undermines digital investigations, data on climate news use by age and imprisoned journalists in China. Click to read, subscribe and share ⬇️ www.linkedin.com/pulse/how-ai...

🚨 Postdoc in Comparative Politics/Public Opinion (2 years) We’re hiring a 100% Postdoc at the University of Greifswald. ✨ What makes this job special: Two full years to focus on research (no teaching, no admin overload) embedded into an International Research Training Group

How AI-generated prose diverges from human writing and why it matters reutersinstitute.politics.ox.ac.uk/news/how-ai-...

Did you know that from tomorrow, Qualtrics is offering synthetic panels (AI-generated participants)? Follow me down a rabbit hole I'm calling "doing science is tough and I'm so busy, can't we just make up participants?"

🔥 What if your next “AI health tip” was totally made up — but sounded 100% credible? In her article, Vanda Boštíková explores how even advanced #LLMs like #GPT4o, #Gemini15Pro, #Llama3 and #GrokBeta can be manipulated to produce false health-related content. Read more below 👇

Vibe coding is the new buzzword, but it raises serious concerns, particularly with regard to security-sensitive applications. See arxiv.org/html/2512.03... and medium.com/@michaelhend...

How cheap is it to build a bot army? 🤖 Find out using the first global index to track real-time prices for verifying fake accounts on 500 platforms, from TikTok to Amazon 👇 https://bit.ly/44s9QSu

“While the AI industry claims its models can “think,” “reason,” and “learn,” their supposed achievements rest on marketing hype and stolen intellectual labor. In reality, AI erodes academic freedom, weakens critical reading, and subordinates the pursuit of knowledge to corporate interests.”

I have just got back from the University of Bucharest, where 21 students from seven EU universities that are part of CIVIS did a one-week training course on disinformation. Read more about the programme "Online fake news and disinformation: recognize and verify": civis.eu/en/learn/civ...

I love when colleagues share amazing stories with me! Thanks @olivierklein.bsky.social They Droned Back - Young journalists expose Russian-linked vessels circling off the Dutch and German coast www.digitaldigging.org/p/they-drone...

AI Can Write Your Code. It Can’t Do Your Job. https://terriblesoftware.org/2025/12/11/ai-can-write-your-code-it-cant-do-your-job

If you’re in the business of facts and accuracy, you can forget entry into the United States. Apparently they want none of that. Makes sense for an emerging autocracy www.poynter.org/news-release...

The first depressing news I received in my inbox today was this: "I was forced to use AI until the day I was laid off." Copywriters reveal how AI has decimated their industry - via @bcmerchant.bsky.social

From ‘News aggregation’ to more than 430 other entries, this encyclopedia is packed with discoveries, thanks to the incredible work of its editors. Huge congratulations! www.researchgate.net/publication/...

After 3 years and 5,322 emails (and counting) our Encyclopedia of Political Communication is finally out at @elgarpublishing.bsky.social 📚 Three volumes ⭐ 431 entries 🎓 581 wonderful authors from across the world A very short 🧵

Interesting, and I would argue refreshing, job advert from the Economist for a newsroom expert

'Artificial intelligence is still in its infancy. It's a bit like the Wild West at the moment.' (A self-proclaimed AI expert) 'AI is great. It can tell you what meals you can make with the leftovers in your fridge.' (The same expert). Journalists should always double-check the 'experts'.

The Ultimate OSINT Guide: Essential Tools for Phone Number Investigation by eRRor Global Verification and Lookup Tools Data Breach Analysis Carrier, SIM, and Porting Intelligence Phone Number Intelligence & Caller ID Social Media and Username medium.com/@eRRoR_/the-...

“TikTok and other social media platforms are hosting AI-generated deepfake videos of doctors whose words have been manipulated to help sell supplements and spread health misinformation.” www.theguardian.com/society/2025...

The AI Hype Index: The people can’t get enough of AI slop www.technologyreview.com/2025/11/26/1...

Large language models often lie and cheat. We can’t stop that—but we can make them own up. www.technologyreview.com/2025/12/03/1...

Most discussions about “misinformation” treat the problem as if people simply don’t have good enough critical-thinking skills. It sounds intuitive, but it’s not entirely accurate. The problem isn’t just individual ignorance, it’s the structure of the information environment people are placed into.

X blocks EU Commission’s advertising account after €120 million fine

"Even if we see a market correction, the technology will not magically disappear." www.niemanlab.org/2025/12/the-... by @felixsimon.bsky.social

Hallucinated references are a serious matter. One of the most time-consuming tasks faced by peer reviewers and academic teachers is verifying the existence of academic citations. This tool can help with that: citely.ai Please feel free to share any other reliable and accessible solutions.

How it started (quote post) How it's going (linked article) www.theguardian.com/technology/2...

The Belgian government has officially banned the use of DeepSeek on all devices belonging to the federal administration. This preventive measure follows a security assessment which highlighted risks to data confidentiality. www.vrt.be/vrtnws/en/20...

The Digital Omnibus: Deregulation Dressed as Innovation - The EU's sweeping data and AI package loosens safeguards for workers while promising competitiveness gains that will flow mainly to US tech giants. www.socialeurope.eu/the-digital-...

Présenté comme une simplification technique neutre destinée à "libérer la puissance des données" [le digital omnibus] est un exercice de dérégulation qui mettra à l’écart les objectifs sociaux et en matière de travail et affaiblira les protections des travailleurs www.socialeurope.eu/the-digital-...

🚨 New in Nature+Science!🚨 AI chatbots can shift voter attitudes on candidates & policies, often by 10+pp 🔹Exps in US Canada Poland & UK 🔹More “facts”→more persuasion (not psych tricks) 🔹Increasing persuasiveness reduces "fact" accuracy 🔹Right-leaning bots=more inaccurate

Artificial intelligence chatbots are very good at changing peoples’ political opinions, according to a study published Thursday, and are particularly persuasive when they use inaccurate information www.science.org/doi/10.1126/...

Today the @towcenter.bsky.social is releasing a tracker (built by my talented colleague @aisvarya17.bsky.social) that monitors developments between news publishers and AI companies—including lawsuits, deals, and grants—based on publicly available information. tow.cjr.org/ai-deals-law...

Also look at the #DamascusDossier, published some hours ago by @icij.org. "ICIJ, @recherche.ndr.de and 24 media partners in 20 countries spent over eight months organizing and analyzing the documents". @tijd.be @lesoir.be @yle.naamio.social @lemonde.fr @irishtimes.com @svtnyheter.bsky.social

All journalists/adjacent will be next.

AI transcription and translation in journalism: evidence that underrepresented languages consistently underperform in AI models cnti.org/article/ai-t...

I was in London last month for the first in-person JournalismAI Festival hosted by @polislse.bsky.social and the Google News Initiative. I wrote up a few of the highlights for @niemanlab.org. www.niemanlab.org/2025/12/news...