📑 New ETUI technical brief on #AlgorithmicManagement just out ⚖️ A legal commentary from @aida.follows.technology on a recent judgement from the Court of Justice of the European Union which clarifies key provisions under the General Data Protection Regulation #GDPR Read it here 🔗 etui.org/Z9A

‼️"How weakening #duediligence EU rules became a ‘betrayal’ for #workers", important @euobserver.com coverage of #Omnibus I with testimonies from @agathesmyth.bsky.social @aida.follows.technology & many more‼️ #DecentWorkfForAll #StopSocialDumping #Fairness #SocialEurope bsky.app/profile/euob...

𝗡𝘂𝗺é𝗿𝗶𝗾𝘂𝗲 𝗲𝗻 𝗰𝗼𝗺𝗺𝘂𝗻[𝘀] 𝟮𝟬𝟮𝟱 𝗠𝗮𝘀𝘁𝗲𝗿𝗰𝗹𝗮𝘀𝘀. Together with Vincent Mandinaud, ANACT, we discussed how to act upon the growing digital 𝘀𝘂𝗿𝘃𝗲𝗶𝗹𝗹𝗮𝗻𝗰𝗲 at work numerique-en-communs.fr

📽️Link to the recording of the @etui.bsky.social 𝗔𝗜 𝗧𝗮𝗹𝗸 which presented the results of the testing of a specific AI tool in the education sector in Germany www.etui.org/events/chall...

📢 I am hosting today's 𝗔𝗜 𝗧𝗮𝗹𝗸 at European Trade Union Institute (ETUI) 𝗖𝗵𝗮𝗹𝗹𝗲𝗻𝗴𝗲𝘀 𝗼𝗳 𝗘𝗱𝗧𝗲𝗰𝗵 & 𝘁𝗵𝗲 𝗥𝗶𝘀𝗸𝘀 𝗼𝗳 𝗧𝗲𝗰𝗵𝗻𝗼𝘀𝗼𝗹𝘂𝘁𝗶𝗼𝗻𝗶𝘀𝗺 𝗶𝗻 𝗘𝗱𝘂𝗰𝗮𝘁𝗶𝗼𝗻, featuring Julie Lüpkes (@University of Oldenburg) & Marte Henningsen (Maastricht University). 🗓 Join at 14hr00CET ➡️ Registration: :https://lnkd.in/eGXHqEQX

Is Europe investing strategically in the development of 𝗱𝗮𝘁𝗮 𝗰𝗲𝗻𝘁𝗿𝗲𝘀? What are the key material and labour challenges involved? I discuss them in this article ⏬

⁉️What are the concrete & measurable benefits 𝗔𝗜 𝗳𝗼𝗿 𝘄𝗼𝗿𝗸𝗲𝗿𝘀? I am encouraging trade unions to develop a strong strategic plan on AI, one that does not extract workers data but instead enhances their agency

@aida.follows.technology reminds us that AI isn’t neutral. Without transparency and regulation, it can deepen power imbalances at work and within societies. Workers need a say - and unions must be at the table from day one. @etui.bsky.social

The rise of data-extractive tools in education makes it crucial for educators and unions to take an active role in shaping AI governance

At the Hamad Bin Khalifa University, just before presenting the book 𝐀𝐫𝐭𝐢𝐟𝐢𝐜𝐢𝐚𝐥 𝐢𝐧𝐭𝐞𝐥𝐥𝐢𝐠𝐞𝐧𝐜𝐞, 𝐥𝐚𝐛𝐨𝐮𝐫 𝐚𝐧𝐝 𝐬𝐨𝐜𝐢𝐞𝐭𝐲 published by @etui.bsky.social A huge appreciation to all the authors who contributed their insights & research to this volume

𝐏𝐨𝐥𝐢𝐜𝐲 𝐫𝐨𝐮𝐧𝐝𝐭𝐚𝐛𝐥𝐞 𝐨𝐧 𝐀𝐈 𝐥𝐢𝐚𝐛𝐢𝐥𝐢𝐭𝐲 Who bears responsibility when AI causes harm, damage, or loss? Is it the software developer, intermediary, deployer, end-user, regulator, or some other entity? We exchanged a diversity of views in this interesting panel at the 𝐇𝐚𝐦𝐚𝐝 𝐁𝐢𝐧 𝐊𝐡𝐚𝐥𝐢𝐟𝐚 𝐔𝐧𝐢𝐯𝐞𝐫𝐬𝐢𝐭𝐲

As part of @kuleuvenuniversity.bsky.social 600th anniversary, I’m proud to contribute to 𝐀𝐈 𝐃𝐢𝐚𝐥𝐨𝐠𝐮𝐞𝐬 with my essay 𝐆𝐨𝐯𝐞𝐫𝐧𝐢𝐧𝐠 𝐀𝐈 𝐚𝐭 𝐖𝐨𝐫𝐤. I argue for anticipatory foresight and real-time reporting mechanisms so all actors can adopt a preventive oversight approach to AI.

Opinion piece on innovation in the EU

My latest article in Social Europe: The EC argues that Europe must focus on innovation, with regulatory simplification at the centre. This reflects a profound misunderstanding of what innovation actually needs: not regulatory chaos, but a clear one www.socialeurope.eu/ctrlaltdelet...

Join me tomorrow at 🕑2pm CET to discuss with Anders Hulum, University of Chicage about generative AI and the labour market AI chatbots are everywhere in workplaces but how are they changing our jobs, wages, time or bargaining dynamics? Register ⏬ and tune in to hear the evidence from 🇩🇰

📆 One for your post-summer agendas: the first #AItalk of the season! 🔗 etui.org/ZRp

Join me tomorrow at 2pm CET to discuss with Anders Hulum, University of Chicage about generative AI and the labour market. AI chatbots are everywhere in workplaces but how are they changing our jobs, wages, time or bargaining dynamics? Tune in to hear the evidence Register here below ⏬

My latest piece: Protecting data in the EU means protecting an independent EDPS Independence is the foundation of trust in Data Protection Everyone talks about AI, but who protects our data in the EU? An authority that must stay independent My analysis👇🏼 www.computerweekly.com/opinion/Prot...

Granmalty AI grader assesses learning outcomes and automates grading decisions: it claims to predict your grade before submission. It fits perfectly under the Annex III of the AI ACT ans under the profiling definition. www.grammarly.com/ai-agents/ai...

AI is designed to automate: to get rid of workers

The DeepSeek case is interesting: 1) AI is determined by the access to the best chips 2) China could succeed in producing competitive chips 3) EU's position in AI hardware seems more vulnerable Does the EU needs to really accelerate the chips' production? Any views? www.ft.com/content/eb98...

Why do millions of people use Chatgpt? Does giving instructions in plain language through questions or search-like formats mean we're accessing AI intelligence? www.pewresearch.org/short-reads/...

We might be starting the era of AI as "human compatible" (Stuart Russell's book) The most striking news regarding the new GPT5 is (1) Open AI dropped the intelligence hook and focuses on user experience, connections, emotions: “so it’s more like your AI.” www.theatlantic.com/technology/a...

A new era promises to transcend human enhancement: Marketed as a "no die protocol" with "Become a Don't Die Citizen" as a motto, this movement envisions defeating death through optimization & tech dontdie.com

The EU’s AI Act enters its 2nd enforcement phase with rules for providers of general‑purpose AI models (GPAI) and the establishment of full governance structures at EU and national levels. www.itpro.com/business/pol...

I have been working on surveillance for some time and the the overall issues are still poorly discussed at work. The commodification of our personal data and society's entrapment in surveillance systems is a major concern of our digitised lives. Here another study www.which.co.uk/policy-and-i...

Soft law for AI. The Code of Practice is a voluntary framework for AI providers to align with the EU AI Act before August 2, 2025, elements: 1. Transparency requirements 2. Copyright compliance 3. Safety and Security 4.Technical mitigations 5. Independent external assessment 6. Incident reporting

Sam Altman warned at a Federal Reserve conference that entire job categories, notably customer service, could vanish entirely as AI becomes faster & more accurate "Some areas, again, I think just like totally, totally gone" Any projections for new jobs? www.theguardian.com/technology/2...

The artificial intelligence ... Google blames the model for “becoming” oversensitive. But the model didn’t make itself, guys. But how would a model “become” anything? It’s software techcrunch.com/2024/02/23/e...

Very interesting the EC's data disclosure template for GPAI models. I understand that there are many questions but I think that this is the right to trace content used in AI training which is essential and was missing in the AI ACT

Imposing AI use by default: AI features are extremely easy to trigger, but they are also sometimes enabled by default or deployed before users are allowed to choose to deactivate them

Under whose rules does digital governance operate: those of the EU, or those set by tech? Despite being actively involved in lobbying the EV and having access to drafts of the Code, Meta has refused to it. A calculated move to shape the regulatory process! www.politico.eu/article/meta...

Agent to automate our digital lives ChatGPT interacting with computer, open files, etc. But it is geographically restricted: until OpenAI completes regulatory compliance? 'you are always in control' www.theguardian.com/technology/2...

No more deep fakes? Denmark proposes granting people copyright over their digital likeness to prevent misuse Will define unauthorised “very realistic digital representations of a person, including their appearance and voice” hackread.com/denmark-ai-c...

Data protection law at work should work The EDPB states that it will strengthen dialogue with stakeholders: this should also entails dialogue with trade unions & workers, who do need to be consulted and be supported in how best to exercise data rights at work www.edpb.europa.eu/news/news/20...

No to simplification of EU digital laws Weakening the AI Act would undermine accountability, the protection of society and workers and bring legal uncertainty I have co-signed the letter led by @cdteu.bsky.social 👇🏼

⏰ Today, CDT Europe, together with 51 experts, academics & civil society organisations, sent an open letter to the European Commission to express our concerns regarding the forthcoming Digital Simplification package, which could include revisiting the AI Act. 👇🏻 Read the full letter on our website:

A copyright of our own body? Do we need this to be respected ? Danish copyright law would, theoretically, give people in Denmark the right to demand that online platforms remove content if it is shared without consent. www.theguardian.com/technology/2...

AI governance between voluntary Codes of Practice and Consultations aimed at simplifying the AI Act: a delicate dance where everyone is invited to the table. Do not ask who is actually setting the rules ...

The transfer of user data by DeepSeek to China is unlawful. DeepSeek has not been able to convincingly prove to my authority that data from German users

An update on the US AI governance ⭐An AI action plan to reduce regulatory barriers might be released. www.reuters.com/legal/govern...

AI regulation strategic pause? It's a pivotal summer. While the AI Act stands as a global regulatory benchmark, several issues sparked intense calls for a strategic pause ⏸️ The deepfake transparency, the Code of Practice, & the high risk systems www.reuters.com/technology/t...

Meta's $29 billion fundraising for AI infrastructure shows the significant capital for AI development. This has implications for the EU's competitiveness goal and may influence how regulators adapt innovation policies or Omnibus Packages with the AI Act's requirements www.ft.com/content/aff1...

The complex reality beneath the AI surface In the rush to automate, are we able to take time to reflect upon the promised efficiencies and the less-attractive consequences? related to discrimination, surveillance and sustainability? irlpodcast.org/season6/epis...

Is this an AI high-risk use case? At first sight it might not be : avoids remote biometric identification; does not use emotion recognition. But, it can qualify as high-risk if used in worker automated monitoring and algorithmic management www.businessinsider.com/veesion-ai-t...

The EC seeks clarification about the high-risk use cases under the AI Act. This could inform potential Guidelines. This consultation may not address cross-cutting risks that do not fit into existing categories www.riskinfo.ai/post/ai-insi...

AI and disappearing professions Professional translators are in direct and fierce competition with the machine. A concrete example of dehumanisation of work

From toys to software OpenAI expanding its reach, from education, universities to toys. We might see GenAI in Mattel's toys soon. OpenAI continues surprising us new forms of everything in this world, even 'play'. I'm curious, how can the company innovate 'play'? techcrunch.com/2025/06/12/o...

Political tensions in governing AI in the EU The AI Act is facing delays due to 'operational' and political challenges. Some argue that the AI Act is currently “not fit for purpose”

Adopting AI in your workplace? What for exactly? Between the market pull and the adoption reality of AI, managers are underestimating the impact of AI on the workforce www.forbes.com/sites/bernar...