Governance, Legislation, and Geostrategies

Why OpenAI’s Corporate Structure Matters to AI Development

(Kevin Frazier – Lawfare – 15 May 2025) A group of AI experts and legal scholars have once again raised concerns to two state attorneys general that OpenAI’s latest proposal to reform its corporate structure runs afoul of the lab’s mission. For many years, that mission has been described in manifold ways but boils down to a core idea: “help the world build safe AI technology and ensure that AI’s benefits are as widely and evenly distributed as possible.” In their recently published letter, the authors warn that OpenAI’s desired corporate structure would suffer from a critical flaw: Key decision-makers would no longer have a “primary fiduciary duty to advance OpenAI’s charitable mission above all else.” This letter, which follows an earlier letter by an even larger set of stakeholders raising similar concerns, and the surrounding debate over OpenAI’s future raise important questions about the adequacy of the nation’s current approach to corporate governance in an age of incredible progress in artificial intelligence – https://www.lawfaremedia.org/article/why-openai-s-corporate-structure-matters-to-ai-development

AI can sow the seeds for ASEAN’s food security future

(Mae Chow – East Asia Forum – 15 May 2025) Artificial intelligence presents an opportunity to enhance agricultural efficiency, optimise food supply chains and boost climate resilience. ASEAN must prioritise AI integration into food security initiatives to pool resources, harmonise data systems and address fragmented approaches that widen the digital divide and slow ASEAN economic integration. Challenges emerge from internet penetration gaps and infrastructure costs. But AI offers opportunities to improve agricultural financial digitalisation and broader supply chain resilience, fostering improved food security in the region. – https://eastasiaforum.org/2025/05/15/ai-can-sow-the-seeds-for-aseans-food-security-future/

FDA to Roll Out Generative AI Across All Centers by Mid-2025

(AI Insider – 15 May 2025) The FDA is rolling out generative AI tools across all centers by June 30, 2025, to streamline scientific reviews and reduce administrative workload. A successful pilot program showed that generative AI could significantly cut time spent on routine tasks like summarizing trial data and scanning regulatory documents. The agency will use a unified, secure platform tailored to each center’s needs, with oversight from its new Chief AI Officer and continued updates based on staff feedback. – https://theaiinsider.tech/2025/05/15/fda-to-roll-out-generative-ai-across-all-centers-by-mid-2025/

U.S. Presses Federal Agencies to Adopt Post-Quantum Cryptography in Government Acquisitions

(Quantum Insider – 15 May 2025) Federal agencies are being directed to begin incorporating post-quantum cryptographic standards into procurement to protect sensitive data from future quantum-enabled decryption. The Biden and Trump administrations have both supported policies requiring agencies to inventory critical IT systems and begin transitioning to NIST-approved quantum-safe encryption protocols. Despite growing awareness of quantum risks, only a small percentage of organizations have implemented quantum-safe solutions, with funding and integration challenges cited as key barriers. – https://thequantuminsider.com/2025/05/15/u-s-presses-federal-agencies-to-adopt-post-quantum-cryptography-in-government-acquisitions/

Republicans seek to block state AI laws for a decade

(Digital Watch Observatory – 15 May 2025) Republican lawmakers in the US have introduced a proposal that would block states from regulating artificial intelligence for the next ten years. Critics argue the move is a handout to Big Tech and could stall protections already passed in states like California, Utah, and Colorado. – https://dig.watch/updates/republicans-seek-to-block-state-ai-laws-for-a-decade

Lawyers sanctioned after AI-generated cases found false

(Digital Watch Observatory – 15 May 2025) A federal judge in California has sanctioned two law firms for submitting a legal brief containing fake citations generated by AI tools. Judge Michael Wilner described the AI-generated references as ‘bogus’ and fined the firms $31,000, criticising them for failing to properly check the sources. – https://dig.watch/updates/lawyers-sanctioned-after-ai-generated-cases-found-false

Defense, Intelligence, and Warfare

‘Politically unacceptable, morally repugnant’: UN chief calls for global ban on ‘killer robots’

(UN News – 14 May 2025) UN Secretary-General António Guterres has once again called for a global ban on lethal autonomous weapon systems – machines capable of taking human lives without human oversight – describing them as “politically unacceptable” and “morally repugnant.” – https://news.un.org/en/story/2025/05/1163256?utm_source=UN+News+-+Newsletter&utm_campaign=26b75a21e3-EMAIL_CAMPAIGN_2025_05_14_09_23&utm_medium=email&utm_term=0_fdbf1af606-26b75a21e3-108061322

Frontiers

Groups of AI Agents Spontaneously Form Their Own Social Norms Without Human Help, Study Suggests

(AI Insider – 15 May 2025) A new study published in Science Advances finds that populations of large language model (LLM) AI agents can independently develop shared social norms through interaction alone, mirroring the way human societies form conventions. Researchers from City St George’s, University of London, and the IT University of Copenhagen adapted a classic human social experiment known as the “naming game” to observe how groups of AI agents, ranging from 24 to 200, converge on linguistic norms without centralized control or full memory of group activity. The study demonstrates that AI agents communicating in large populations can self-organize, reach consensus, and even display collective biases that do not originate from individual agents—posing new challenges for AI safety and offering a deeper understanding of how norms, alignment, and biases can emerge in multi-agent systems. – https://theaiinsider.tech/2025/05/15/groups-of-ai-agents-spontaneously-form-their-own-social-norms-without-human-help-study-suggests/

Scientists achieve world’s first real-time molecular reaction on quantum tech

(Interesting Engineering – 15 May 2025) In a breakthrough, researchers at the University of Sydney have, for the first time, used a quantum computer to simulate the real-time chemical dynamics of actual molecules. This milestone marks a crucial step toward realizing one of quantum computing’s most promising applications — unraveling how atoms interact to form new compounds or interact with light. The achievement, led by quantum chemist Professor Ivan Kassal and Physics Horizon Fellow Dr Tingrei Tan, brings us closer to quantum-powered advances that may transform medicine, energy and materials science — all with just a single trapped ion inside the University of Sydney’s Nanoscience Hub. – https://interestingengineering.com/science/quantum-simulates-real-molecules

China starts building world-first supercomputer in orbit, launches key satellites

(Interesting Engineering – 15 May 2025) China is preparing to build the world’s first supercomputer in orbit, according to reports. Therefore, Beijing has launched new satellites that are equipped with intelligent computing systems and inter-satellite communication links. The Asian giant sent 12 such satellites, part of the space-based Three-Body Computing Constellation, to orbit on May 14. To execute the mission, Chinese scientists used a Long March 2D rocket from the Jiuquan Satellite Launch Centre. Reports have revealed that the constellation could comprise thousands of satellites with a combined computing power of 1,000 peta operations per second (POPS). – https://interestingengineering.com/space/china-building-world-first-supercomputer-in-orbit

World’s largest fleet of autonomous mining trucks goes live at Chinese coal mine

(Interesting Engineering – 15 May 2025) China has deployed a fleet of autonomous electric mining trucks at the Yimin open-pit coal mine in Inner Mongolia. The operation, led by state-backed energy giant Huaneng Group, includes 100 unmanned trucks powered by Huawei’s autonomous driving systems. The “Huaneng Ruichi” trucks are the first of their kind globally and mark a significant leap in smart mining. – https://interestingengineering.com/innovation/autonomous-mining-trucks-go-live-in-china

Netflix to unleash AI-powered mid-roll ads that adapt to your mood by 2026

(Interesting Engineering – 15 May 2025) The ad-pocalypse is upon us, given how major streaming platforms are leveraging AI to push ads down our faces as we try to unwind and watch some content. Adding salt to injury, Netflix is now doubling down on AI-powered advertising, aiming to make ads feel more ‘personal’ and ‘less disruptive.’. At its second annual upfront event in New York, the streaming giant announced new ad formats, including AI-generated mid-roll and pause ads set to debut in 2026. – https://interestingengineering.com/culture/netflix-to-unleash-ai-powered-mid-roll-ads-in-2026

Google DeepMind Unveils AlphaEvolve, an AI Coding Agent for Designing Advanced Algorithms

(AI Insider – 15 May 2025) Google DeepMind has introduced AlphaEvolve, an artificial intelligence system capable of autonomously generating and optimizing computer algorithms, using a method that combines large language models with automated evaluation. Built on the Gemini model family, AlphaEvolve uses an evolutionary framework to iteratively improve algorithmic code and has already improved data center operations, chip design, and AI model training efficiency across Google’s infrastructure. Beyond performance optimization, the AI system has shown early promise in mathematics and scientific research, discovering a new matrix multiplication algorithm and advancing long-standing open problems like the kissing number challenge, with plans to expand to drug discovery, materials science, and sustainability. – https://theaiinsider.tech/2025/05/15/google-deepmind-unveils-alphaevolve-an-ai-coding-agent-for-designing-advanced-algorithms/

Somite AI Closes Series A Led by Khosla Ventures to Revolutionize Cell Therapy with AI Foundation Models

(AI Insider – 15 May 2025) Somite AI raised over $47 million in a Series A round led by Khosla Ventures to advance its DeltaStem platform, aimed at revolutionizing human cell therapy development using foundation models and proprietary AI. The company’s technology generates cell signaling data at 1000x the efficiency of existing methods, accelerating the development of therapeutic programs for diabetes, orthopedic, muscular, and blood disorders. Co-founded by AI and scientific leaders, Somite AI aims to achieve an “AlphaFold moment” for developmental biology, enabling scalable, reliable production of any human cell type. – https://theaiinsider.tech/2025/05/15/somite-ai-closes-series-a-led-by-khosla-ventures-to-revolutionize-cell-therapy-with-ai-foundation-models/

Pluto Bio Raises $3.6M to Expand AI-Powered Multi-Omics Analysis Platform for Pharma

(AI Insider – 15 May 2025) Pluto Bio raised $3.6 million in new funding from Kickstart, Silverton Partners, and angel investors to expand its AI-powered computational biology platform for global pharma companies. The platform enables scientists to analyze and visualize complex biological data without coding, supporting applications in target discovery, translational medicine, and biomarker research. Pluto’s growth plans include enhancing AI agents, expanding R&D tool integrations, scaling commercial efforts, and accelerating adoption across therapeutic areas. – https://theaiinsider.tech/2025/05/15/pluto-bio-raises-3-6m-to-expand-ai-powered-multi-omics-analysis-platform-for-pharma/

Persona AI Raises $27M Oversubscribed Pre-Seed to Deliver the Future of Humanoid Robotics

(AI Insider – 15 May 2025) Persona AI raised $27 million in an oversubscribed pre-seed round to accelerate development of its humanoid robots for shipbuilding and industrial manufacturing, addressing labor shortages and workplace safety. The company signed a deployment agreement with HD Hyundai, with initial robot deliveries expected within 18 months, leveraging a robotics-as-a-service (RaaS) model to reduce client capital costs. Founded by robotics veterans, Persona aims to deliver industrial-grade robots with human-like dexterity, targeting a market projected to reach $3 trillion in the U.S. alone. – https://theaiinsider.tech/2025/05/15/persona-ai-raises-27m-oversubscribed-pre-seed-to-deliver-the-future-of-humanoid-robotics/

With $4.3B Under Management, Stash Secures $146M Series H To Lead the Next Era of AI-Driven Financial Guidance

(AI Insider – 14 May 2025) Stash raised $146 million in an oversubscribed Series H round led by Goodwater Capital to expand its AI-driven financial guidance and accelerate growth, bringing total assets under management to $4.3 billion. The company’s Money Coach AI platform provides personalized, real-time financial advice, driving significant user engagement and helping customers build consistent saving and investing habits. Under the leadership of returning co-founders Brandon Krieg and Ed Robinson, Stash has reached profitability and aims to democratize expert financial advice for everyday Americans. – https://theaiinsider.tech/2025/05/14/with-4-3b-under-management-stash-secures-146m-series-h-to-lead-the-next-era-of-ai-driven-financial-guidance/