🚨Results are in: #EU Parliament has upheld its ban on most forms of public facial recognition in the #AIAct, as well as biometric categorisation of sensitive characteristics & emotion recognition in key sectors! This is a landmark step against biometric mass surveillance! But..

🚨🚨EU AI Act Plenary Vote🚨🚨 For anyone wondering "what does this mean for ChatGPT/generative AI", today's vote has essentially confirmed the text voted at committee level in May So the measures for 'foundation models' have not been changed since that last vote

Gestern gab es die langersehnte Abstimmung des Europäischen Parlamentes zum #AIAct. Damit heißt es: It´s trilogues time!

Diese Woche war der AI Act im Europaparlament, nächste Woche gibts eine Sonderausgabe unseres #Erlesenes-Newsletters dazu 🎉📚Interesse an Lektüreempfehlungen rund um das Thema? Dann gehts hier zum Abo: reframetech.de/newsletter/ Die Blogparade gibts auf reframetech.de

Es 👏 ist 👏 keine 👏 Planwirtschaft 👏 wenn👏 man 👏 einen👏 Plan 👏 hat 👏 um 👏 Klimaziele 👏 einzuhalten.👏 Jesus Maria sind wir hier in der 8. Klasse.

Oft fragen mich Leute mit einer "Gotcha-Haltung" nach Atomkraft. Ich habe nur einen pragmatischen Einwand, der aber gilt absolut gerade. Sie ist einfach viel zu teuer, wie dieser Chart von Azeem Azhar zeigt:

🚨 SCOOP: OpenAI lobbied the E.U. to weaken forthcoming AI regulation, even as in public it calls for stronger AI guardrails, documents obtained by TIME show. We're publishing the key document in full alongside my story 👇 time.com/6288245/openai…

The AI Act is not groundbreaking,but at its best it provides useful transparency & documentation obligations. The loopholes in the Council (Art 6(3)) & Parliament (Art 6(2)) texts threaten to allow companies to opt out of the very basic requirements at will. This must be changed

#BeyondAICollective On Monday we (the Beyond AI Collective) had the opportunity to discuss discrimination in the context of AI-systems with @HannahRusc Michael Puntschuh and Nina Bewig (LADS Berlin) and many more participants. Thx! #AIethics

1/2🌍💡#KünstlicheIntelligenz & #Nachhaltigkeit - it's complicated! Die neue #Erlesenes-Ausgabe beschäftigt sich deswegen ausschließlich mit diesem Thema, u.a. mit Artikeln zum Wasserverbrauch & Emissionen großer Sprachmodelle u.v.m. Lest mal rein📚: reframetech.de/2023/07/06/erl…

#AIAct trilogue In this debate about prohibitions, civil society is not ‘extreme’ & the Parliament position is not something ‘extreme’ that needs to be toned down Our position on bans is backed up by the EDPS EDPB Volker Türk UN OHCHR & numerous experts

The Council wants AI-powered racist profiling The Council wants dystopian AI lie detectors for police & borders The Council wants to use bullshit racist AI to predict “future criminality” This stuff is extremist, racist, dystopian & has no place in our society #NoBansNoDeal

Human rights advocates, academics, EU data protection bodies, UN High Commissioner for Human Rights, EU anti-discrimination and national human rights institutions are calling for bans, while EU states come with "but criminals abuse tech, so should we" arguments to #AIAct debates.

🚨 #AIAct negotiators must not trade away our human rights. @Amnesty, with civil society & digital human rights experts, is calling on EU institutions to unequivocally reject attempts to legalise dangerous, discriminatory police AI. edri.org/our-work/ngos-…

Maybe my auto-translation isn't correctly translating. But is this a complaint from a public official that people are trying to pierce through a notoriously non-transparant public policymaking process?

The thing is, law enforcement representatives have always had access, industry has always had access, but thankfully this time the public has known about & been able to prevent disastrous compromises ‘Confidentiality’ has always worked well for powerful actors with access

🚨AI Act Deal🚨 There's been a deal, at least on most of the text The issue is that we don't know what was sacrificed The PR is a big danger They will say "we banned the most dangerous uses & added robust safeguards for other uses" But details matter x.com/EUCouncilPress…

Wise words from Daniel Leufer. Whatever we see on biometrics/LE, it's going to be disappointing because the Council demanded 'compromise' on things that should never be compromised. The question is not whether the final result will be good, but whether it's the least bad...

An illustration of the issue we now face in evaluating this #AIAct deal Dragoș Tudorache says "we have a ban on biometric categorisation, on predictive policing, on emotion recognition" etc But we do not know what the loopholes, the exceptions are, so those bans could be useless

The EU's landmark #AIAct was agreed at minutes to midnight yesterday. Whilst lawmakers celebrate, human rights groups are sceptical. As someone who's spent years calling on the EU to #BanBiometricMassSurveillance via Reclaim Your Face, here's my take on where they landed: