Done another update to the Tech Policy Jobs (Europe) sheet, including two from us @algorithmwatch.org ! There's a lot of new jobs flying around, it's hard to keep track, so please do add your own new jobs if you're hiring. docs.google.com/document/d/1...

Forgot to post this here - latest blog piece, on academic / non-academic collaborations www.oliver-marsh.com/academic-non... (If you subscribe you don't need to rely on me remembering to post them here too ;) )

Very happy to be a signatory to this letter drafted by @hutko.bsky.social countering Trumpian misinformation on DSA. Please read thread. I also note Martin's kind offer to provide DSA training to US politicians at 1% cost 🤣

1/4 Yday there was a hearing on "Europe’s Threat to American Speech and Innovation" in the US Congress. It's the latest in misleading attacks on Europe's ability to have its own laws. See this open letter signed by 40 NGOs inc. @algorithmwatch.org about that: algorithmwatch.org/en/open-lett...

Non-consensual sexualization tools #NSTs are apps and websites that generate sexualized images of real people. Often called #NudifyApps, they don’t only create nudity but can also include non-consensual “undressing” via images in underwear or swimsuits: algorithmwatch.org/en/stop-nudi...

Using my current laptop to search for a new laptop feels morally wrong somehow :/

Part 2 of my thoughts on reaching a more "pro-democracy internet". In it I attempt to categorise & compare 4 models - from New Coffee Houses to Better Broadcasters - across a few different axes such as "expected participation" and "theory of change". www.oliver-marsh.com/a-better-int...

Google Trends is getting an API, very useful development for internet research developers.google.com/search/blog/...

👋 For a 5th time, we are looking for new Algorithmic Accountability Reporting fellows. Apply now if you have research ideas concerning the relation between Artificial Intelligence and power and its consequences. Application deadline: 15 September 2025 23:59 CET algorithmwatch.org/en/open-call...

Last few days to apply for our Technology & Democracy role (until Mon 14)

Last week, the Delegated Act on data access under the DSA was published. If it works effectively, it allows researchers to access to internal data of very large online platforms and search engines. The clue here is the 'if'. More: 👉 algorithmwatch.org/en/dsa-deleg...

On the off-chance anyone is trying to look at my personal website, it's currently offline while I migrate to a proper newsletter system. For now, if you want to be added to the subscription list, let me know. Posts coming up on pro-social design and how to predict the future [of political culture].

Today, we will submit expert input to the EU on systemic risks stemming from online platforms and search engines. We argue risks come not just from technology, but also (1) the attitudes of companies and (2) lack of transparency around enforcement. Read our full response here: t1p.de/uoc2y

Looking forward to this tomorrow!

It was great working with @markscott.bsky.social on this German election research, which featured in the The Washington Post yesterday. Summary 🧵 below. Full report on @algorithmwatch.bsky.social website (auch auf Deutsch). www.washingtonpost.com/world/2025/0...

Could everyone please stop posting about going to RightsCon I'm getting FOMO

Olga and Riddle Russia have been doing important analysis for many years, but are now at genuine financial risk after recent developments. If you can help personally, or know of organisations who can help them get back to stability, please do say. I can put you in touch. ridl.io/donate/

Useful long thread summary from Joanna on AI Action Summit panels and talks

Ranking of tweets with the most community notes: #1 x.com/realDonaldTr...

Any EU legal eagles got thoughts on the question below? @mathver.bsky.social ?

Just realised Thursday was my 3 year anniversary in Berlin. Pains of migration and bureaucracy aside, it's been an absolutely great decision.

My latest newsletter is some further reflection on Meta and online speech, plus Rupert Murdoch. (You can subscribe to the newsletter at the top right corner of the page). www.oliver-marsh.com/post/sideway...

This from @klonick.bsky.social and @reckless.bsky.social was, as per, great. But q: does "the state still has police and jails" apply to the EU/DSA? Eg if a platform isn't following transparency provisions & refuses fines, noone goes to jail, right...? (Durov was got under French law, iirc)

If you are offering or looking for a job in tech policy, in Europe (inc UK): I host a doc of interesting-looking jobs. I am bad at regularly updating it but have just done a refresh. If you have an ad out for a relevant job, feel free to add it as a comment. docs.google.com/document/d/1...

I am also getting blocked on #democrats, #republicans, #liberal, #liberals, #conservatives... But not #democrat #republican or #conservative

TL;DR - the "Meta fact-checking was negligible" story may be more complicated, from DSA data. This (excellent, must-read) post by @jonathanstray.bsky.social cites @columjournreview.bsky.social and @popular.info low numbers of fact-checks in the USA (somewhere between 10 and 14 per day, in 2020)

I've been exploring how X Community Notes Data might be used for social media monitoring. Found some possibly interesting approaches: in-progress but wanted to share where it's at. When more robust I can share code etc. Please do share with the social media monitoring folks in your life!

anyone want to join me in a Community Notes rabbithole? notes rated "helpful" are autotweeted (for different languages) so you can see them all communitynotes.x.com/guide/en/und... Skimming recents: EN, lots of notes on tweets abt LA fires, German ones on AfD plus other crime/rightwing stuff

Seeing a lot of (interesting and useful!) debate about fact-checkers vs. community notes. But seems to me it would likely be best if Meta did *both*...?

I really, really don't want to do this, but let's do a bit (just a LITTLE BIT) of fact checking about what Zuck said on the Joe Rogan podcast, because this is something I not only followed closely, but also spent time *talking to the actual people at Meta who dealt with this* which WAS NOT ZUCK.

Interesting points from Laura Edelman about the “more personalized approach to political content” part of the Meta announcement, which she sees as a return to political content ranking a la 2016-2020. x.com/LauraEdelson...

Mark Zuckerberg’s decision to gut moderation & fact-checking on Instagram, Facebook and Threads shows he prioritizes pleasing #Trump over societal well-being. Moving moderation to Texas? A transparent ploy for Republican approval, not ‘de-biasing.’ algorithmwatch.org/en/zuckerber...

Point 6 particularly relevant for DSA folks, Musk is not the only issue here

Huh, @poorlydrawnlines.bsky.social just published this a moment ago

German politics watchers: this "we said we wouldn't coalition with far-right but turns out alternatives are too complicated" in Austria is what I worry about for next #Bundestagswahl in Germany, but I've been told this is still unlikely from CDU. Any views? www.zeit.de/politik/ausl...

I found this useful and also interesting for those of us engaged in critical approaches to LLMs, particularly some of the points around getting better and worse for the environment

If you were going to write an opinion piece with a prominent coauthor on a somewhat-wonky critique of big tech assault on European democracy, where would you most want to put it? It's still of general interest about how democratic legitimacy works, aimed at the tech-interested general audience.

Nick Clegg leaving Meta, seems they really are going all in on making republicans comfortable with them www.ft.com/content/e5fe...

The BBC needs to have a debate about whether Musk's outbursts should keep leading the news. This isn't about censoring or suppressing his views; it's about whether to amplify them. Most Brits aren't on X, they don't follow Musk & not every raving from a foreign plutocrat should set the news agenda

"Banking, by Kafka" is a great title for episode 1 of our (@algorithmwatch.bsky.social's) first-ever podcast - on automated banking, #scoring & the consequences for all of us. I can really recommend it if you'd like to understand better the phenomenon called #de-banking. Let me know what you think!

Happy new year all. I wrote some reflections on the last year and also the year ahead, around the themes of elections, impunity, networks, and random wisdom www.oliver-marsh.com/post/sideway...

Good paper ⬇️. But I still - genuine q, beyond this paper - wonder abt value of focussing on info that is "false or misleading", rather than harmful irrespective of truth value. And D/Misinfo does get lots of focus, eg EU Code of Practice on Disinfo is much higher profile than C.O.P. on Hate Speech.

@algorithmwatch.bsky.social ending 2024 with a new LLMs & elections report. Also a test of DSA 40.4 Data Access - we even got data! Key points in thread 👇 (while travelling so maybe installments). Was quite a process, so lots to say! @clarahelming.bsky.social algorithmwatch.org/en/llms_stat...

This is a good example of why generative AI opt-outs don’t work: this is my music, but I have no way of opting it out of AI training. I composed this piece, but I didn’t make the recording and I don’t control the website where it’s hosted. www.boosey.com/audio-clip/I... 1/4

Mathias isn't the only person I know who I was stunned to learn dedicated part of his life to documenting directly the horror in Syria. Both are now working in making us safer in the EU. Thank you.

Hey, everyone. As you hopefully already know, we're Kickstarting an absolutely amazing card game where you get to run a social media network. Since it's a game about social media, and this is (looks around) a cool social media site, we had a request... www.kickstarter.com/projects/mma...

Initial quick response (in German) to the first #DSA risk and audit reports, contact me if you'd like to chat about them further (in English or German) algorithmwatch.org/de/dsa-risik...

Delighted to announce the prizewinner for "most ironic place to be stuck on a stationary train"

The article suggests a lot of this money would be spent on ads. Political ad reform in the UK is urgently needed.

Initial quick response (in German) to the first #DSA risk and audit reports, contact me if you'd like to chat about them further (in English or German) algorithmwatch.org/de/dsa-risik...