Just because X is using an image generator to do it and takes requests through an automated interface doesn't change the fundamentals of what is happening - X is producing and distributing CSAM by request In law, it must make no difference: those are criminal offences commited *by X*

Nudify apps have been around for ages, but it's wild to see a mainstream platform like X effectively become a revenge porn machine, generating AND publishing non-consensual intimate images of women

I think this feels particularly shocking because no matter how much you stretch the law this has absolutely no basis of legality whatsoever. No terrorists, no "threat" requiring anticipatory self defence. Nothing.

☕️📖 Nothing like spending a cozy morning curled up under a blanket with a cup of tea and a book on my lap to keep me company while I endlessly scroll my phone

NEW: Hackers stole more than $2.7 billion in crypto last year, according to multiple blockchain monitoring companies. North Korea alone stole 2 of those billions, thanks to the jackpot of the Bybit hack, which netted them $1.4 billion. techcrunch.com/2025/12/23/h...

This is the biggest data harm case in the history of the UK and it just got bigger. Because of a bug in Fujitsu system hundreds of postmasters were sued, jailed and bankrupted. Now it's revealed THE POST OFFICE KNEW IT WAS A BUG AND NOT A FEATURE. And they continued destroying people's lives.

If this is already the normal procedure everywhere, i.e. a google scholar import, universities’ library catalogues will be useless tomorrow.

People who aspire to be academics: I wish to live a life of the mind Actual academics:

NEW: judge-alone trials would save just 2% of time in the Crown Court. In total, we estimate reforms to jury trials would save <10% of court time. Great write-up from @reporterrwright.ft.com here www.ft.com/content/8a9f...

Yesterday, Ofcom at last published the small services FAQs document on its website. www.ofcom.org.uk/online-safet...

Grading and googling hallucinated citations, as one does nowadays, and now that LLMs have been around for a while, I've discovered new horrors: hallucinated journals are now appearing in Google Scholar with dozens of citations bc so many people are citing these fake things

The amount of money Meta made from Chinese scam ads last year is more than double Reddit's entire business. Have fun on Instagram today!

In 2024 en 2025 zijn er veel rapporten verschenen over AI en criminaliteit. De Politie, Europol, het NCSC, UNODC en inlichtingendiensten hebben er allemaal wel iets over gepubliceerd.

An academic disliked an Oxford Very Short Introduction (145 pages) in his field so much that he wrote a 200 page book review attacking it. www.pierre-legrand.com/ewExternalFi...

The Australian social media ban seems destined for fines against tech companies for failure to take "reasonable steps" to remove children. The difficulty is that to really identify all children--including children using their parents' accounts--you need draconian steps. 1/

Portugal updates cybercrime law to exempt security researchers www.bleepingcomputer.com/news/securit...

Predatory journals + LLM generated papers =

In all the jury kerfuffle what isn't being discussed and should be is magistrates. The mixed bench proposed by Leveson to replace juries has been quietly dropped in favour of a single judge. And mags sentencing powers are going up again to 18 months (+up to 2 yrs) www.theguardian.com/law/2025/dec...

Where is the scientific evidence? www-bbc-com.cdn.ampproject.org/c/s/www.bbc....

New #CalledToTheBar! Imogen Saunders chats with editors from the American Journal of International Law, Oxford Journal of Legal Studies, and Australian Yearbook of International Law about publishing, peer review, & what makes a great article. Essential ECR listening! soundcloud.com/calledtotheb...

Relevant to law enforcement use of AI-based systems in countries everywhere.

If the figures in this article are anything like accurate, #scam centres may now be the largest industry in Laos, Cambodia and Myanmar... to have got this large without the significant involvement of senior government officials, army, police etc. is inconceivable.

This was a very interesting conversation and now I really want to know more about the bespoke MLATs!

So chosen because of the way it will prompt those online to say "actually, that is two words". qed.

EU Member States agree on negotiation position on CSA material online - will now negotiate with the European Parliament www.consilium.europa.eu/en/press/pre...

the moral of the story here is, after committing a serious crime, immediately google • how to prove innocence falsely accused • normal to feel deep empathy for others? • recipes that take four hours to make starting at 9pm • how to report legitimate windfall to irs reddit • best orphan charities

Methodology

This headline is really misleading. “Chat control” is dead in that the new text has removed detection orders which would have required scanning private E2EE chats. This preserves the status quo, which allows service providers to voluntarily scan their services for CSAM as many already do.

British and Australian chemists have discovered a powerful new antibiotic called pre-methylenomycin C lactone, hiding in a well-known soil bacterium. This molecule kills drug-resistant bacteria without triggering resistance. buff.ly/YlXaONo #ShareGoodNewsToo

I have two PhDs and honestly managing to successfully connect to the university printing service is probably my greatest accomplishment

One of the most interesting recent privacy developments is the deployment of big two-hop IP blinding VPNs by companies like Apple and Google. These systems are designed to ensure that even those companies can’t link web requests to IP addresses.

There’s a BBC story about the Sara Sharif murder review and the child welfare advocate said it should be illegal for parents to physically punish a child: “You can’t hit an adult, why should you be able to hit a child?” What a straightforward way to cut through the noise and make a legal argument.

Studies of abusers on dark web forums show that while mainstream AI models can produce CSAM, it’s not very realistic, so they use LoRAs to make it so. This testing won’t really do much because abusers aren’t using vanilla models. www.bbc.co.uk/news/article...

welcome to bluesky. if the word "mapquest" means anything to you and/or you used it as a verb, you're in the right place

Book Launch! Please join us on Wednesday 19th November @qmul.bsky.social at Queen Mary to hear the wonderful Alan Norrie discuss his latest book. Tix available online or in person - see below! Please repost.

My latest Lawfare piece is out! I look at the recent #latombe case and the impact it will have on #EU / #US #dataprotection and #data #flows. #law #academia #lawfare #cjeu

Academics in Assyria in the 7th c BC complain that admin is preventing them from doing research and teaching

Hackers have launched five cyberattacks against Britain's drinking water suppliers since the beginning of last year, according to reports filed with the drinking water watchdog and partially disclosed to Recorded Future News under freedom of information laws.

Did you happen to leave your @londonreviewbookshop.co.uk purchases on the Metropolitan line yesterday? Did one of your friends do this and send an anguished message to the group chat? If so, Darren has just done you a very good turn. 👇

"Meanwhile, the regulations to bring in a requirement on providers to report CSAM content to the National Crime Agency have been revoked; they were due to come into force on 3 November and, at the time of writing, there are no further details available as to why they have been reversed." > Curious.

An LLM-produced essay is tangible proof that a student doesn’t care, and yet responding to it properly requires hour upon hour of careful work. It’s asymmetrical and overwhelming.

My latest post analyses the potential criminal liability in the UK of AI companies in light of the August report from Anthropic that its generative AI tool Claude has been used by hackers to commit cybercrimes. TL;DR: s3A CMA is a problem. open.substack.com/pub/issuesin...

Always love a good analysis of CMA s3A but what's really interesting is if AI can do all these things successfully for criminals why can they not seem to for lawful markets! I'm not seeing such intelligent price discrimination...

My latest post analyses the potential criminal liability in the UK of AI companies in light of the August report from Anthropic that its generative AI tool Claude has been used by hackers to commit cybercrimes. TL;DR: s3A CMA is a problem. open.substack.com/pub/issuesin...

More evidence that the UK Government not only forgets that NI exists, it also forgets that the Common Travel Area exists. This is just a remarkable story of benefits being withdrawn from hundreds of families for using the CTA amid a performative benefit crackdown: share.google/NSWTG5JNYMrz...

This is nuts. Too much money is going to specific grants about AI and STEM, and all the others need to fight for scraps. We need more funding for social sciences and the arts immediately.

🫠

In celebration of #InternationalOpenAccessWeek, be sure to check out the latest ICLQ articles via FirstView @cambup-law.cambridge.org, covering everything from #ExtraterritorialJurisdiction in digital investigations to the ICJ's #Palestine #AdvisoryOpinion. www.cambridge.org/core/journal...

Any suggestions for things to do in (or near) Leiden with a 4 year old?