Turing laureates headline ACM's inaugural AI summit in Atlanta. ACM's inaugural AI Leadership Summit opens in Atlanta with Turing laureates Yann LeCun and Andrew Barto, plus UN and UNESCO leaders. What the program signals. Key takeaways. * 1ACM's first AI Leadership Summit runs its main program August 31 to September 2 at the Hyatt Regency Atlanta, with a Doctoral Consortium held August 30 at Georgia Tech. * 2The opening track pairs Turing laureates Yann LeCun and Andrew Barto with roboticist Rodney Brooks, three researchers who have each publicly questioned the field's dominant assumptions about scaling language models. * 3Governance and agentic AI are programmed on separate days with almost no overlap in personnel, a structural signal that policy and engineering communities are still converging rather than merged. The Association for Computing Machinery has spent nearly eight decades running the conferences where computing gets argued out. Today it opens its first summit dedicated entirely to artificial intelligence, and the guest list reads less like a trade show and more like a standing committee on the future of the field. The inaugural ACM AI Leadership Summit runs its main program August 31 through September 2 at the Hyatt Regency Atlanta, preceded by an invitation-only Doctoral Consortium at Georgia Tech on August 30. Two ACM A.M. Turing Award laureates - Yann LeCun and Andrew Barto - headline the opening day, joined by iRobot founder Rodney Brooks, the United Nations Secretary-General's Envoy on Technology, a former UNESCO Assistant Director-General, and chief scientists from Microsoft, Adobe, IBM, and The Walt Disney Studios. What makes the program worth reading closely is not the star count. It is the sequencing. ACM has put its frontier-model skeptics on stage first, quarantined governance in its own track with no engineers on the panel, and scheduled agentic AI a full two days later. For executives trying to read where the professional computing establishment actually stands on AI in 2026, the schedule itself is the story. Day one is three skeptics in a row. The Monday morning track is titled Rethinking the Future: Frontier Models and Technologies for a New Era of AI, and it opens with back-to-back talks from Yann LeCun, Andrew Barto, and Rodney Brooks, followed by a panel called The Next Decade of AI moderated by Georgia Tech's Mark Riedl. Each of the three has spent the past several years pushing against the industry's center of gravity. LeCun has argued publicly and repeatedly that large language models are a dead end for genuine machine intelligence, and he has put considerable capital behind that position. Barto, who shared the 2024 Turing Award with Richard Sutton for the conceptual and algorithmic foundations of reinforcement learning, represents a research lineage that predates the transformer era by decades and never depended on it. Brooks has been attacking the field's assumptions about embodiment and symbolic reasoning since the 1980s, while building robot families - Roomba, PackBot, Baxter, Carter - that actually shipped. LeCun's presence carries a specific commercial subtext. He left Meta at the end of 2025 after more than a decade leading its Fundamental AI Research lab, and now appears on the ACM program as Executive Chairman of AMI Labs. In March 2026, AMI closed a $1.03 billion seed round at a $3.5 billion pre-money valuation, co-led by Cathay Innovation, Greycroft, Hiro Capital, HV Capital, and Bezos Expeditions, with Nvidia, Temasek, and Samsung among the participants - reported by TechCrunch as the largest prelaunch AI raise in European history. The company is building world models on LeCun's Joint Embedding Predictive Architecture, aimed at manufacturing, robotics, aerospace, and biomedical applications rather than chat. A keynote arguing that the industry's dominant architecture is a detour lands differently when the speaker has just raised a billion dollars to prove it. The summit by the numbers. Figures drawn from ACM's official summit materials and press coverage of the participants. Main program, Aug 31 - Sep 2 ACM AI Leadership Summit registration page Concurrent SIG tracks on September 2 ACM published program Standard non-member registration after June 30 ACM AI Leadership Summit registration page Student ACM member registration after June 30 ACM AI Leadership Summit registration page Turing Award prize, funded by Google ACM, March 2025 AMI Labs seed round, LeCun's new venture TechCrunch, March 2026 Why ACM built this event. ACM framed the summit as a cross-sector convening rather than a technical conference, positioning it alongside its governance and education work rather than its research proceedings. " Artificial intelligence is transforming every dimension of human knowledge, creativity, and collaboration. The ACM AI Leadership Summit will be a milestone event where the global computing community taps into the excitement of the moment while exploring the AI era from a whole range of perspectives. - Elisa Bertino, ACM President-Elect and co-chair of the Summit organizing committee How the three days are structured. The main program is organized as thematic day-blocks rather than parallel research tracks, with the exception of Wednesday's cross-disciplinary sessions. Track titles and speaker assignments below follow the program published on the summit site as of late July 2026 and may shift on the day. | Day | Tracks | Featured participants | What to watch for | | Sun, Aug 30 (Georgia Tech) | Doctoral Consortium - invitation only | Summit headliners acting as mentors | Early-career pipeline; not part of the public program | | Mon, Aug 31 | Rethinking the Future; AI and Scientific Discovery; AI Governance | Yann LeCun, Andrew Barto, Rodney Brooks, Amanda Randles, Amandeep Singh Gill | Whether the architecture debate gets a real airing or a polite one | | Tue, Sep 1 | AI and the Creative Arts; AI and the Workforce | Aaron Hertzmann (Adobe), Markus Gross (Disney/ETH Zurich), Jaime Teevan (Microsoft), Ed Chi (Google DeepMind), Ellen Zegura (NSF) | Authorship and labor framing from the companies actually shipping the tools | | Wed, Sep 2 | Agentic AI; A Cross-Disciplinary Perspective (six concurrent SIG tracks) | Ece Kamar (Microsoft Research), Dinesh Verma (IBM), Ahmed E. Hassan, Claire Le Goues (CMU), Armando Solar-Lezama (MIT) | The trust question for AI-written software, asked by the people building the verification tools | Governance and agents are in different rooms - On purpose. The most revealing thing about the program is what does not overlap. Monday's AI Governance track features a talk by Amandeep Singh Gill, the UN Secretary-General's Envoy on Technology, followed by a panel moderated by Dame Wendy Hall of the University of Southampton. The panelists are Virginia Dignum of Umeå University, Gabriela Ramos formerly of UNESCO, Francesca Rossi of IBM, Philip Thigo from the Office of the President of Kenya, and Yi Zeng of Renmin University of China. It is a policy panel staffed by policy people. Wednesday's Agentic AI session is three talks - Ahmed E. Hassan of Queen's University, Ece Kamar of Microsoft Research, and Dinesh Verma of IBM - framed by organizers around autonomous systems that plan, reason, and act, with safety, control, and trust presented as engineering consequences rather than regulatory categories. No panel, no policymakers. That separation answers a question a lot of enterprise buyers have been asking implicitly: is agentic AI a capability term or a governance term? On this program it is clearly a capability term, and governance got its own day two slots earlier with an almost entirely different cast. The practical implication for anyone building an internal AI policy is that the two vocabularies have not converged yet, and a compliance framework written from governance literature will not map cleanly onto what deployment engineers are actually shipping. Wednesday afternoon offers the sharpest technical session of the week: a track titled around whether software written by AI can be trusted, moderated by Claire Le Goues of Carnegie Mellon, with Armando Solar-Lezama of MIT, Lin Tan of Purdue, Erik Meijer, Margaret-Anne Storey, and Srikanth Yoginath of Oak Ridge. Program repair, program synthesis, and ML-for-correctness researchers debating verification is the closest thing the summit has to a hard engineering verdict. What business and Technology leaders should take from the program. * Treat the LeCun keynote as a capital-allocation signal, not just a research opinion - world-model architectures now have over a billion dollars of seed funding behind them and named industrial targets in robotics, aerospace, and biomedical work. * Assume your agentic AI vendor conversations and your AI compliance conversations are using different definitions of the same words, and force a shared glossary internally before either one drives a purchase. * Watch the Wednesday software-trust track for the state of the art in verifying AI-generated code - the answer there matters more to engineering leadership than any keynote. * Note the workforce panels are staffed by Accenture, Cisco, Microsoft, and Google DeepMind rather than by economists, which tells you the framing will be product-and-deployment led. * Georgia Tech, Emory, Georgia State, Kennesaw State, and Morehouse are all listed among the supporting institutions - a useful map of Atlanta's AI talent pipeline for anyone recruiting in the Southeast. * Registration ran $300 for ACM members and $400 for non-members after the June 30 early-bird deadline, with student rates at $200 and $250 - modest by industry-conference standards, which reflects the academic-society framing. Frequently asked questions. #ACM AI Leadership Summit #Yann LeCun #Andrew Barto #AI governance #agentic AI #Turing Award #Atlanta tech #world models #event-coverage #ACM AI Leadership Summit 2026

10 European startups shaping our robotic future. August 18, 2026 Europe's robotics and physical AI ecosystem is expanding beyond traditional industrial automation to include humanoid robots, autonomous defence and security systems, software-defined manufacturing, and AI models designed to understand and act in the physical world. This list highlights some of Europe's most promising robotics startups founded between 2024 and 2026. Together, they demonstrate the growing convergence of artificial intelligence, robotics, autonomy and advanced engineering. While not all of these companies are pure-play robotics startups, each is developing technology that enables machines, autonomous systems or industrial infrastructure to operate more intelligently in real-world environments. Several have also attracted significant early-stage investment, reflecting growing investor interest in the technologies underpinning the next generation of physical automation. Advanced Machine Intelligence. Founded in Paris in 2026, Advanced Machine Intelligence develops AI systems that build abstract representations of physical and digital environments from sensor data, allowing intelligent systems to reason, predict outcomes and plan actions. Its work has potential applications across robotics, industrial automation, wearables and healthcare. Advanced Machine Intelligence raised approximately €890 million at a €3 billion pre-money valuation in one of Europe's largest-ever Seed rounds in March this year, Cambridge Aerospace. Operating from Cambridge in the United Kingdom, Cambridge Aerospace was founded in 2024 and develops air-defence technology for detecting and intercepting aerial threats. Its work includes drone and missile interceptors, radar systems and rocket propulsion technology, alongside manufacturing capabilities for its defence systems. Cambridge Aerospace has raised more than $630 million to date, which includes a Series C in August of this year at a €2.94 billion valuation. Flexion Robotics. Zurich is home to Flexion Robotics, a robotics company founded in 2024. It is developing an AI-powered autonomy software stack for humanoid robots, combining reinforcement learning, language-based reasoning, vision-language-action models and whole-body control. The company trains its systems in high-fidelity simulations to help robots learn complex tasks while reducing reliance on human demonstrations. Flexion Robotics has raised approximately €49.6 million to date. Generative Bionics. Founded in 2025 in Genoa, Italy, Generative Bionics develops humanoid robots and Physical AI systems for industrial and service environments. Its approach brings together robotics, artificial intelligence and engineering to support safe and reliable operation in real-world settings. Generative Bionics has raised €70 million in funding so far. Harmattan AI. Also emerging from Paris' growing defence technology ecosystem, Harmattan AI was founded in 2024 and develops AI-powered autonomous systems for military applications. The company's technology covers unmanned vehicles, sensor integration, mission planning and real-time decision-making, with systems designed for operations across land, air and maritime environments. Harmattan AI has raised approximately €209.2 million in funding including a Series B in January of this year. Hive Robots. In Kongens Lyngby, Denmark, Hive Robots is building humanoid robot workforce solutions for businesses. Founded in 2025, the company offers proof-of-concept programmes, robot-as-a-service models, software development and vertically integrated deployments. Hive targets sectors including logistics, retail, hospitality and healthcare across Scandinavia and Northern Europe. Humanoid. With its headquarters in London, Humanoid is developing general-purpose humanoid robots for industrial and commercial environments. Its technology combines proprietary humanoid hardware with its KinetIQ AI and software stack, initially targeting applications across manufacturing, logistics and warehousing. Humanoid raised €133 million in Series A funding in July 2026, reaching a €1.1 billion post-money valuation. Isembard. London-headquartered Isembard is building software-powered factories for high-precision manufacturing across aerospace, defence and energy. Founded in 2024, the company operates modular manufacturing units equipped with machine tools and robotics, producing precision components from materials including metals, plastics and composites. In March of this year, the company raised €43 million ($50 million) in Series A funding, less than 12 months after its Seed round. Rollo Robotics. Founded in Tallinn in 2025, Rollo Robotics develops autonomous monowheel security robots that use gyroscopic stabilisation technology. The company combines robotics, sensing and automation to support continuous security and monitoring applications. Rollo Robotics has so far raised €3.7 million in pre-Seed funding. Stateful Robotics. Emerging from Oxford's robotics research ecosystem, Stateful Robotics was founded in 2025 and develops a hardware-agnostic mission-control platform for mobile robots operating in dynamic environments. Its technology has been used with quadruped robots carrying out hazardous inspections, mobile service robots working in security and care settings, and underwater robots collecting environmental data. Stateful Robotics has raised €4.1 million. Summary. Taken together, these companies illustrate how Europe's robotics sector is developing beyond traditional industrial automation. Their approaches span foundational Physical AI, humanoid platforms, autonomous defence systems, security robots and software for coordinating robotic fleets, with applications across manufacturing, logistics, defence, healthcare and other physical environments. Funding levels vary considerably, but the scale of recent rounds is notable. Advanced Machine Intelligence secured close to €890 million in 2026, while Cambridge Aerospace raised more than €430 million across two rounds during the year. Harmattan AI's $200 million Series B and Humanoid's €133 million Series A further point to substantial capital moving into the intersection of AI, robotics and autonomous systems.

AdaJEPA - Yann LeCun's Lab builds a World Model that keeps learning while it acts. A world model that stops learning the moment it is deployed will plan against the wrong imagined future the moment reality drifts. AdaJEPA - new work from NYU's Agentic Learning AI Lab and AMI Labs by Ying Wang, Oumayma Bounou, Yann LeCun and Mengye Ren - makes the world model adapt inside the planning loop itself: at every model-predictive-control step the agent plans, acts, observes the resulting transition, and uses that transition as a self-supervised training signal before the next replan. The adaptation is deliberately lightweight - one gradient step per replan, a five-transition replay buffer, updates restricted to the final layers - yet it consistently improves planning under distribution shift: unseen object shapes, corrupted observations, changed physics and held-out maze layouts. In low-data regimes it can more than double a frozen model's success rate and even beat frozen models trained on far more data. Vivax unpack the plan-act-adapt-replan loop, the results, and why a clinical world model - deployed into hospitals where the data distribution never stops moving - should be built to keep learning too. What NYU released. NYU's Agentic Learning AI Lab and AMI Labs have released AdaJEPA: An Adaptive Latent World Model - new work by Ying Wang, Oumayma Bounou, Yann LeCun and Mengye Ren, with the paper on arXiv and code on GitHub. The one-line idea: instead of freezing a world model after training and hoping it generalizes, AdaJEPA adapts the model during deployment, inside the planning loop itself. Every action the agent executes produces an observed transition - what it saw, what it did, what happened next - and AdaJEPA uses that transition as a self-supervised training signal before the next replan. Plan, act, adapt, replan. The frozen world model problem. The problem it attacks is one every world-model system inherits. Latent world models make planning from high-dimensional observations tractable by predicting future states in a compact representation space - but they are usually kept frozen after training. When the deployed environment drifts from the training distribution - different object shapes, corrupted camera input, changed physics - the model's imagined futures go subtly wrong, and model-predictive control then optimizes actions for the wrong imagined future. The plan is coherent; the world it was planned for no longer exists. How AdaJEPA works. AdaJEPA starts from a pretrained JEPA world model - a sensory encoder, an action encoder, and a latent predictor - and plans with model-predictive control in latent space: roll the predictor forward over candidate action sequences and pick the one that lands closest to the goal's representation. The new part comes after each step. The observed transition goes into a small online buffer, and the model takes a gradient step on the latent prediction error before the next replan. The recipe is deliberately minimal: one gradient step per replan, a replay buffer of just five recent transitions, and updates restricted to the final layers of the visual encoder and predictor. This is test-time adaptation woven directly into the control loop - cheap enough to run at every step. What the results show. The results track the promise. In-distribution, adaptation helps when the frozen model is suboptimal and does no harm when it is already near-optimal. Under distribution shift, the gains are consistent: on PushObj with held-out object shapes, on PushT with blurred, noisy, darkened or recolored observations, and on PointMaze with altered physics and unseen layouts, each observed transition recalibrates the model before the next replan, improving success rates and pulling trajectories closer to shortest paths. The data-scaling result is the sharpest: when offline training data is limited, test-time adaptation can more than double the frozen model's success on seen shapes - and even outperform frozen models trained on far more data. The takeaway for world models. The visualizations make the mechanism legible: decoded rollouts from the adapted model still reconstruct sensibly on unseen shapes and corrupted views while retaining training-domain structure, which suggests adaptation works by exploiting shared latent structure and recalibrating predictions rather than wandering off the learned manifold. The authors' takeaway is bigger than the benchmarks: world models should continue learning during deployment rather than remain frozen after training. Perception and planning get more resilient when the model treats every step it takes as evidence about the world it is actually in. Why this matters for clinical world models. For medicine, this is not a robotics curiosity - it is the core problem restated. A clinical world model is deployed into permanent distribution shift: new scanners and assays, new hospitals and populations, evolving protocols and drug availability. A model frozen at training time starts aging the day it ships. AdaJEPA is a blueprint for the alternative - grounded models that recalibrate from the transitions they actually observe, under tight, auditable update budgets (one step, five transitions, final layers only) rather than open-ended retraining. That restraint matters doubly in healthcare, where every update must remain safe and traceable. It also extends the JEPA lineage Vivax has been tracking - from the original architecture through V-JEPA 2, Neuro-JEPA and IQ-JEPA - one more step toward Yann LeCun's vision of models that learn how the world works, and keep learning it.