Legal in the Loop

Something is shifting in how companies are hiring for AI governance roles. Six months ago, most job descriptions for AI governance leads were written by legal teams. The requirements read like compliance checklists: policy writing, regulatory mapping, risk documentation. Now I am seeing a different kind of JD. They want someone who can build and run the operational layer. Not just document risk. Actually manage it in production. That is a different hire. It requires programme management instincts, not just legal or policy expertise. The roles are also moving up in seniority faster than expected. What was a mid-level specialist role a year ago is now appearing at senior manager and director level. My read: companies that launched AI products in 2024 and 2025 are now hitting governance problems they didn't plan for. And they are realising that a policy document doesn't fix an operational failure. The people who can sit at the intersection of ops, compliance, and AI systems are going to be in demand for the next few years. The market is not saturated yet. #AIGovernance #AIPolicy #ResponsibleAI #ProgramManagement #FutureOfWork

What does the EU AI Act’s draft Article 50 Guidelines mean in practice? The recent European Commission’s draft Article 50 Guidelines provide further clarity on how transparency obligations will work in reality from August 2026. Our experts Dr. Simon Hembt and Oliver Belitz unpack the key developments, focusing on what matters for organisations navigating compliance. From the technical requirements around marking and detection of AI-generated content to the evolving deepfake regime, the draft guidance provides a clearer picture of expectations for both providers and deployers. With consultation open until 3 June 2026 and enforcement on the horizon, the draft guidance signals a shift from high-level principles to operational reality, with significant implications across the AI value chain. Interested in what this means for your organisation? Read the full article here: https://2bird.ly/4fu4bBp

BREAKING NEWS: EU publishes guidelines for High-Risk AI Systems TODAY The European Commission has just dropped its eagerly anticipated guidelines on high-risk AI system compliance. This comes just days after EU leaders reached a political agreement to delay the compliance date for high-risk AI systems to 2 December 2027. The guidelines published today are focused on the classification of high-risk AI systems. In particular, they support organisations in answering the following questions: ➡️ Is our AI system high-risk? ➡️ What are the general principles that make an AI system high-risk? ➡️ What is the 'filter' to exempt AI systems from being high-risk? ➡️ How can we demonstrate and document that our AI system is not high-risk (i.e., because it performs a narrow procedural task)? There is also a detailed breakdown of every type of high-risk AI system listed in Annex III. For each category, there is an overview of cross-cutting issues, as well as a deep-dive on every category and sub-category of high-risk AI system. This gets as granular as AI systems used in the context of road traffic management, evaluating learning outcomes, and evaluating credit scores. For each category and sub-category, tangible examples of high-risk AI systems are provided. This will make it much harder for organisations to claim that their AI system is not high-risk (if it is listed in the guidelines). These guidelines are vital, as they provide clarity on exactly what is and is not high-risk and how this will be determined by regulators and courts over time. However, the new guidelines are also 167 pages long! No one said understanding the EU AI Act was easy. These are draft guidelines that are open for comment. My book, Fundamentals of AI Governance, features a dedicated and up-to-date chapter on the EU AI Act. Pre-order it today in advance of the September release. I can promise that it will be more visually pleasing than any official guidance could ever dream of!

As global discussions on AI governance intensify, the shift from non-binding ethical guidelines towards binding international legal frameworks is becoming increasingly necessary. Key developments such as the 𝗖𝗼𝘂𝗻𝗰𝗶𝗹 𝗼𝗳 𝗘𝘂𝗿𝗼𝗽𝗲’𝘀 𝗙𝗿𝗮𝗺𝗲𝘄𝗼𝗿𝗸 𝗖𝗼𝗻𝘃𝗲𝗻𝘁𝗶𝗼𝗻 𝗼𝗻 𝗔𝗜 𝗮𝗻𝗱 𝗛𝘂𝗺𝗮𝗻 𝗥𝗶𝗴𝗵𝘁𝘀 (𝟮𝟬𝟮𝟱), the 𝗨𝗡 𝗚𝗹𝗼𝗯𝗮𝗹 𝗗𝗶𝗴𝗶𝘁𝗮𝗹 𝗖𝗼𝗺𝗽𝗮𝗰𝘁 (𝟮𝟬𝟮𝟰), and the draft 𝗨𝗡 𝗖𝗼𝗻𝘃𝗲𝗻𝘁𝗶𝗼𝗻 𝗼𝗻 𝗔𝗜, 𝗗𝗮𝘁𝗮, 𝗮𝗻𝗱 𝗛𝘂𝗺𝗮𝗻 𝗥𝗶𝗴𝗵𝘁𝘀 (𝗠𝘂𝗻𝗶𝗰𝗵 𝗖𝗼𝗻𝘃𝗲𝗻𝘁𝗶𝗼𝗻, 𝟮𝟬𝟮𝟰) are shaping a rapidly evolving regulatory landscape. Concurrently, initiatives within the European Parliament continue to push the debate towards the possibility of a global AI treaty 🔗 https://bit.ly/4uaYJrM On 𝗧𝗵𝘂𝗿𝘀𝗱𝗮𝘆, 𝗠𝗮𝘆 𝟮𝟭𝘀𝘁, 𝟮𝟬𝟮𝟲 (𝟭:𝟯𝟬–𝟱:𝟬𝟬 𝗣𝗠), a select group of international and interdisciplinary experts will convene at the Technical University of Munich for a roundtable discussion titled: “𝗔𝗜 𝗚𝗼𝘃𝗲𝗿𝗻𝗮𝗻𝗰𝗲 𝗶𝗻 𝗮 𝗙𝗿𝗮𝗴𝗺𝗲𝗻𝘁𝗲𝗱 𝗟𝗮𝗻𝗱𝘀𝗰𝗮𝗽𝗲” The event is jointly organized by: - Technical University of Munich (TUM) – Institute for Ethics in Artificial Intelligence - TUM - Ludwig-Maximilians-Universität München (Ludwig Maximilian University of Munich) - Ludwig Boltzmann Institute of Fundamental and Human Rights - University of Vienna The roundtable will explore: - Convergences and divergences among emerging regulatory frameworks - Regional perspectives from Europe, the Americas, and the Global South - Pathways towards a globally binding framework for AI governance under international law A central outcome will be a strategic report which synthesizes stakeholder perspectives on human rights–based AI governance, with a particular focus on the implications of the proposed UN Convention on AI, Data, and Human Rights. #TUMThinkTank #AIGovernance #AIethics #HumanRights #InternationalLaw #AIpolicy #Munich #TechPolicy Christoph Lütge, Caitlin Corrigan, Alexander Kriebitz, Stefanos Athanasiou, Dr. Camilla Haake, Univ.-Prof. Dr. Michael Lysander Fremuth 

We’re hiring for a lead to shape our approach to structured adversarial safety testing for Bytedance's GenAI products and models. Please consider applying — or share with someone who’d be energized by building new capabilities from the ground up, navigating fast-moving challenges, and doing high-impact work at massive scale (with a little healthy chaos along the way). US-based role: https://lnkd.in/gku7vQVi SG-based role: https://lnkd.in/gtF_W2hR

We’re hiring for a lead to shape our approach to structured adversarial safety testing for Bytedance's GenAI products and models. Please consider applying — or share with someone who’d be energized by building new capabilities from the ground up, navigating fast-moving challenges, and doing high-impact work at massive scale (with a little healthy chaos along the way). US-based role: https://lnkd.in/gku7vQVi SG-based role: https://lnkd.in/gtF_W2hR

The OpenAI Deployment Company just launched with $4 billion and 19 investors — including TPG, Goldman Sachs, McKinsey & Company, and Bain Capital. 2,000+ portfolio companies are about to receive Forward Deployed Engineers who will build AI into their most critical workflows. Not one of those engineers is a legal practice. Not one of them provides attorney-client privilege. Not one of them continously studies the failure patterns that will emerge from these deployments — because that's not their job. Global Data Innovation's sole focus is to study AI failure playing out somewhere else — and build the system to stop it before it reaches you. The deployment machine just arrived. The governance gap it created is this week's most important conversation for every GC, CAIO, Chief Risk Officer, CEO and Board in its path. globaldatainnovation.com #AIGovernance #OpenAI #BoardLeadership #TRUSTFramework #CorporateGovernance

We're hiring a Policy Counsel in EMEA to help shape the future of AI regulation. You'll work with Matt Cagle and me, as well as the broader Anthropic Legal team and our exceptional colleagues in Public Policy. If you're a creative and collaborative lawyer who cares about responsible AI, please apply.