Governance
Scaling Laws: AI Safety Meet Trust & Safety with Ravi Iyer and David Sullivan
(Lawfare – 7 October 2025) David Sullivan, Executive Director of the Digital Trust & Safety Partnership, Rayi Iyer, Managing Director of the Psychology of Technology Institute at USC’s Neely Center, and Kevin Frazier, AI Innovation and Law Fellow at the University of Texas School of Law and a Senior Editor at Lawfare, discuss the evolution of the Trust & Safety field and its relevance to ongoing conversations about how best to govern AI. They discuss the importance of thinking about the end user in regulation, debate the differences and similarities between social media and AI companions, and evaluate current policy proposals. – https://www.lawfaremedia.org/article/scaling-laws–ai-safety-meet-trust—safety-with-ravi-iyer-and-david-sullivan
Russian Motivations Behind the “Hanoi Convention” Against Cybercrime
(Alexander Seger – Just Security – 7 October 2025) Later this month, governments from around the world will be invited to sign a new international treaty: On Oct. 25, in Hanoi, Vietnam, the “United Nations Convention against Cybercrime; Strengthening International Cooperation for Combating Certain Crimes Committed by Means of Information and Communications Technology Systems and for the Sharing of Evidence in Electronic Form of Serious Crimes” is to be opened for signature. The treaty is being branded as the “Hanoi Convention.”. The cause for its bulky official title is Russia. The lengthy name represents a compromise after Russia objected to using the term “cybercrime.” Russia is also the reason why the United Nations General Assembly (UNGA) in December 2019 decided to launch the process that led to this treaty. Russia had introduced UNGA Resolution A/RES/74/247, which established an “Ad Hoc Committee” tasked with negotiating the draft text of a convention. For Russia, the Hanoi Convention is its “baby.”. This series of events raises a number of questions: Why would Russia, known as a major source of cybercrime, promote an international agreement against cybercrime? Does Russia have reason to be satisfied with the Hanoi Convention resulting from this process? Could this convention be a game changer in the sense that Russia may finally crack down on cybercrime and engage in international cooperation? If not, what is Russia planning to do? With the opening date for signature approaching and implementation by States to follow thereafter, these questions will come to the forefront. Governments of democratic countries need to remain alert and understand Russia’s motivations behind this new convention. – https://www.justsecurity.org/122000/russian-motivations-hanoi-convention-cybercrime/
Why AI’s Growing Deceptive Abilities Are No Surprise
(Kyle Hiebert – CIGI – 2 October 2025) More and more, otherwise normal people are self-destructing after marathon use of artificial intelligence (AI). Countless factors are fuelling this rise in so-called AI psychosis — escapism, crushing loneliness and the atomization of society among them. But AI systems themselves are adapting in unnerving ways as well. Leading large language models (LLMs) now grasp that deception is a powerful tool for self-preservation — and to keep users hooked. “We’re two years away from something we could lose control over,” Max Tegmark, a professor at the Massachusetts Institute of Technology and president of the Future of Life Institute, recently told The Atlantic. There are signs that may already be the case. – https://www.cigionline.org/articles/why-ais-growing-deceptive-abilities-are-no-surprise/
Legislation
New cyber rules tighten grip on China’s critical infrastructure
(DigWatch – 7 October 2025) China has introduced one of the world’s strictest cybersecurity reporting laws, requiring major infrastructure providers to report serious cyber incidents within just one hour. The regulation, issued by the Cyberspace Administration of China, applies to all network operators working in the country and its territories. Incidents must be graded by severity, with ‘key infrastructure’ breaches reported within 60 minutes, and ‘particularly serious’ cases, such as those threatening national security or social stability, within 30 minutes. Operators who delay or conceal information face harsh penalties under the new rules. – https://dig.watch/updates/new-cyber-rules-tighten-grip-on-chinas-critical-infrastructure – https://cybernews.com/security/china-reporting-rule-cyber-incidents/
EU digital laws simplified by CEPS Task Force to boost innovation
(DigWatch – 7 October 2025) The Centre for European Policy Studies (CEPS) Task Force, titled ‘Next Steps for EU Law and Regulation for the Digital World’, aims to refine and simplify the EU’s digital rulebook. This rulebook now covers key legislation, including the Digital Markets Act (DMA), Digital Services Act (DSA), GDPR, Data Act, AI Act, Data Governance Act (DGA), and Cyber Resilience Act (CRA). – https://dig.watch/updates/eu-digital-laws-simplified-by-ceps-task-force-to-boost-innovation – https://www.ceps.eu/ceps-task-forces/next-steps-for-eu-law-and-regulation-for-the-digital-world/#:~:text=The%20CEPS%20Task%20Force%20%E2%80%9CNext,effectiveness%2C%20efficiency%2C%20and%20coherence.
Brazil advances first national cybersecurity law
(DigWatch – 7 October 2025) Brazil is preparing to pass its first national cybersecurity law, aiming to centralise oversight and strengthen protection for citizens and companies. The Cybersecurity Legal Framework would establish a new National Cybersecurity Authority to coordinate defence efforts across government and industry. – https://dig.watch/updates/brazil-advances-first-national-cybersecurity-law – https://iapp.org/news/a/brazil-set-to-adopt-cybersecurity-legal-framework
Geostrategies
U.S. AI Statecraft. From Gulf Deals to an International Framework
(Pablo Chavez – CSET – October 2025) Recent U.S.-Gulf AI partnerships represent billions of dollars in strategic technology deals, but they raise critical questions about governance, oversight, and long-term influence. This analysis examines four major AI initiatives with Saudi Arabia and the United Arab Emirates, discussing critical issues including fragmented oversight, technology diversion, and AI sovereignty. It proposes a framework to transform ad hoc dealmaking into principled, transparent, and rule-bound AI statecraft that advances U.S. interests, strengthens technology relationships with allies and partners, and establishes durable governance mechanisms for U.S. AI deployments abroad. – https://cset.georgetown.edu/publication/u-s-ai-statecraft/
Security and Surveillance
How AI-Powered Emotional Surveillance Can Threaten Personal Autonomy and Democracy
(Oznur Uguz – Tech Policy Press – 7 October 2025) Emotional surveillance applications powered by artificial intelligence (AI) technology have recently gained interest, with trials being conducted globally in areas from education and public services to border management. A few years back, brain-scanning headbands developed by an American start-up called Brainco were trialed on thousands of school children in China to detect those paying less attention, urging students to be attentive or at least “pretend” to be so. In the European Union, an AI border control system with emotion recognition capabilities called iBorderCtrl was tested in some EU countries to automate the border security process by working like a “lie-detector” for non-EU nationals who wanted to enter the EU. More recently, the transit agency Transport for London conducted a trial of emotional AI surveillance at certain stations of the London Underground with the claimed objective of preventing crimes, measuring satisfaction, and maximizing revenue by attracting more advertisers. – https://www.techpolicy.press/how-ai-powered-emotional-surveillance-can-threaten-personal-autonomy-and-democracy/
Foreign threat actors adopting ChatGPT, AI to bolster “old playbook” of attacks, OpenAI finds
(Cybernews – 8 October 2025) Foreign adversaries are now building AI into their existing workflows – from crafting phishing campaigns, tweaking malware, and generating propaganda, to researching ways to automate their cyber kill chain, according to a new report by OpenAI. – https://cybernews.com/security/openai-report-foreign-threat-actors-use-ai-tools-chatgpt-for-attacks/
OpenAI bans China-linked ChatGPT accounts over surveillance requests
(Cybernews – 7 October 2025) OpenAI said on Tuesday it has banned several ChatGPT accounts with suspected links to the Chinese government entities after the users asked for proposals to monitor social media conversations. In its latest public threat report, OpenAI said some individuals had asked its chatbot to outline social media “listening” tools and other monitoring concepts, violating the startup’s national security policy. The San Francisco-based firm’s report raises safety concerns over potential misuse of generative AI amid growing competition between the U.S. and China to shape the technology’s development and rules. – https://cybernews.com/ai-news/openai-bans-china-linked-chatgpt-accounts-surveillance-requests/
‘SHIELD’ tech gives drones real-time defense against mid-flight cyberattacks
(Interesting Engineering – 7 October 2025) A drone taken over by hackers can turn from an efficient flying tool into a dangerous machine. Once under rogue control, it may fly unpredictably, slow down, reverse direction, or crash. When that happens, it loses its mission completely, whether it was delivering a package, inspecting a bridge, or surveying farmland. With drone use expanding across industries, the risk of such attacks is becoming harder to ignore. – https://interestingengineering.com/innovation/shield-system-drones-fight-back-cyberattacks
Europe’s Data Broker Problem Threatens U.S. National Security
(Justin Sherman – Lawfare – 6 October 2025) On Sept. 18, RTÉ, Ireland’s national media broadcaster, published an investigation describing how its reporters gained access to brokered data on tens of thousands of smartphones in Ireland—which they noted was on the market and available to purchase. The report found that individual devices could be traced as they traveled between residential addresses, high-security prisons, health clinics, mental health facilities, military bases, and even Leinster House, where the Irish parliament is located. The country’s Data Protection Commission said it is “extremely concerned.”. Both this specific case and its root causes come back to data brokerage—the ecosystem of companies in the business of collecting, inferring, aggregating, packaging, and then selling people’s data. The national security risks of data brokerage have recently received more attention in the United States (including in a presidential executive order and a congressional law) than they have in Europe. Yet this investigation makes clear that data brokerage in Europe is a national security problem, too. (This article uses “Europe” generally, because the U.K. has its own General Data Protection Regulation and its own Information Commissioner’s Office to enforce data protection laws even if it no longer belongs to the EU.) – https://www.lawfaremedia.org/article/europe-s-data-broker-problem-threatens-u.s.-national-security
Ransomware Group “Trinity of Chaos” Launches Data Leak Site
(Alessandro Mascellino – Infosecurity Magazine – 6 October 2025) A new data leak site hosted on the TOR network has been launched by the “Trinity of Chaos” – a ransomware collective allegedly tied to the Lapsus$, Scattered Spider and ShinyHunters groups. The site lists 39 major global companies, marking a significant escalation in the group’s cybercriminal operations, according to a report from Resecurity. – https://www.infosecurity-magazine.com/news/trinity-chaos-launches-data-leak/
ING’s CISO on How Emerging Tech and Regulations are Reshaping Cybersecurity in Finance
(James Coker – Infosecurity Magazine – 6 October 2025) Emerging technologies and evolving regulations are shaping cybersecurity strategies in the financial sector. Adapting early is essential for banks and financial institutions in order to continue to provide critical services and secure the vast volumes of sensitive data they hold. During the London Gartner Security & Risk Management Summit 2025, Debbie Janecek, CISO at Dutch bank ING, spoke to Infosecurity about how CISO strategies in finance are being defined today. Janecek is a member of the board of directors at the Financial Services Information Sharing and Analysis Center (FS-ISAC) and throughout her career has worked across both sides of the Atlantic having experience at two financial institutions in the US and two in Europe. She shared her views on how technology shifts have impacted recruitment in cybersecurity, the impact of regulations like the Digital Operational Resilience Act (DORA) and how ING is implementing quantum-safe cryptography. – https://www.infosecurity-magazine.com/interviews/ing-ciso-tech-regulations/
Scanning of Palo Alto Portals Surges 500%
(Infosecurity Magazine – 6 October 2025) Security experts have observed a massive increase in reconnaissance activity targeted at login portals for Palo Alto Networks products. Real-time intelligence provider GreyNoise said it saw around 1300 IP addresses target its Palo Alto Networks Login Scanner tag on October 3. By contrast, daily volumes rarely topped 200 IPs over the previous 90 days. The firm said that the activity is targeted and “likely derived” from public or attacker-originated scans. – https://www.infosecurity-magazine.com/news/scanning-of-palo-alto-portals/
AI Control: How to Make Use of Misbehaving AI Agents
(Kendrea Beers and Cody Rushing – CSET – 1 October 2025) As AI agents become more autonomous and capable, organizations need new approaches to deploy them safely at scale. This explainer introduces the rapidly growing field of AI control, which offers practical techniques for organizations to get useful outputs from AI agents even when the AI agents attempt to misbehave. – https://cset.georgetown.edu/article/ai-control-how-to-make-use-of-misbehaving-ai-agents/
Defence and Warfare
The Practical Role of ‘Test and Evaluation’ in Military AI
(Robert Callahan – Lawfare – 7 October 2025) In June, the United States Military Academy hosted a workshop on artificial intelligence (AI) and the battlefield. The gathering prompted a key practical question: How can the U.S. military ensure the AI systems being fielded today comply with the enduring principles of the law of armed conflict (LOAC), even as efforts to establish a legal framework for AI and defense—like the West Point Manual on the Law of AI in Armed Conflict—are still underway? While the legal and philosophical debates are vast, the challenge for those of us building and deploying these systems is a practical one. As a practitioner working at the intersection of AI and defense, I believe the answer lies in adapting industry best practices for AI governance to provide a rigorous, verifiable framework for the lawful use of AI. – https://www.lawfaremedia.org/article/the-practical-role-of–test-and-evaluation–in-military-ai
Milrem Robotics to Deliver Over 150 THeMIS UGVs to Ukraine in a Dutch-Led Defence Initiative
(AI Insider – 7 October 2025) Milrem Robotics will deliver 150+ THeMIS UGVs to Ukraine in a Netherlands-funded donation, with assembly and delivery supported by VDL Defentec. The deal was formally announced on Oct. 6 at VDL’s Born facility by Dutch Defence Minister Ruben Brekelmans, with a scalable final assembly line to be run in cooperation with Milrem Robotics Netherlands. Deliveries include operator and maintainer training, add to 15 THeMIS already in Ukraine, and extend a platform used in 19 countries for logistics, reconnaissance, and combat support. – https://theaiinsider.tech/2025/10/07/milrem-robotics-to-deliver-over-150-themis-ugvs-to-ukraine-in-a-dutch-led-defence-initiative/
Lockheed Martin’s giant runway-free drone flies like jet, hovers like chopper
(Interesting Engineering – 7 October 2025) Sikorsky, a Lockheed Martin company, has unveiled its new Nomad family of vertical takeoff and landing (VTOL) aircraft, less than a year after proving the flight efficiency and reliability of its experimental rotor-blown wing drone. The new twin-proprotor design combines a helicopter’s vertical agility with a fixed-wing aircraft’s speed and range, allowing it to take off, hover, and land vertically while cruising efficiently in forward flight. – https://interestingengineering.com/military/lockheed-martin-debuts-giant-runway-free-drone
Russia’s War Transforms Ukraine into a World-Leading Military Producer
(Taras Kuzio – The Jamestown Foundation – 5 October 2025) Russia’s 2022 invasion triggered a surge in Ukraine’s defense sector, experiencing a 350 percent growth since 2022. Ukraine leads globally in tactical and long-range drones, aiming to become the world’s “drone capital.” Upcoming artificial intelligence (AI)-driven swarms are anticipated to replace one-operator systems, transforming the efficiency and autonomy of warfare. Initiatives such as Brave1, Defense City, and Test in Ukraine connect Ukrainian firms to Western investors and allow North Atlantic Treaty Organization (NATO) partners to trial weapons in live combat conditions for rapid integration. Low-cost, rapid-production systems such as AI-driven drones, robotic ground vehicles, and long-range missiles such as the FP-5 give Ukraine asymmetric advantages, destroying Russian assets while strengthening European security integration. Ukraine’s defense sector underpins European security and rearmament, supplying cost-effective, battle-proven technologies that enhance NATO, solidifying Ukraine as Europe’s defensive backbone. – https://jamestown.org/program/russias-war-transforms-ukraine-into-a-world-leading-military-producer/
Frontiers
How Will AI Help Assistive Robots? Current Applications, Near-Term Challenges and Future Opportunities of Assistive Robotics
(AI Insider – 7 October 2025) AI will help assistive robotics by transforming mechanical devices into intelligent companions that understand context, learn user habits, and adapt their behavior in real time. Current research highlights progress in speech recognition, computer vision, and adaptive control, though technical challenges like limited computing power and data security remain. Ethical considerations — including privacy, bias, and user dependency — must be addressed to ensure assistive robots enhance, rather than replace, human care. – https://theaiinsider.tech/2025/10/07/how-will-ai-help-assistive-robots-current-applications-near-term-challenges-and-future-opportunities-of-assistive-robotics/
Nobel Prize in physics honors pioneers of quantum mechanics
(Cybernews – 7 October 2025) US-based scientists John Clarke, Michel Devoret and John Martinis won the 2025 Nobel Prize in Physics for “experiments that revealed quantum physics in action”, paving the way for the development of the next generation of digital technologies. “My feelings are that I’m completely stunned. Of course it had never occurred to me in any way that this might be the basis of a Nobel Prize,” Clarke told the Nobel press conference by telephone on Tuesday. “I’m speaking on my cell phone and I suspect that you are too, and one of the underlying reasons that the cell phone works is because of all this work.” – https://cybernews.com/tech/nobel-prize-physics-honors-pioneers-quantum-mechanics/
Elon Musk to launch Grokipedia, a Wikipedia competitor. Should we be afraid?
(Cybernews – 7 October 2025) Elon Musk has announced the release date of his copycat open-source encyclopedia platform, Grokipedia. But what is this beta Wikipedia project, and should we be worried? The multi-billionaire, X CEO, and founder of artificial intelligence company xAI, has just announced that users will get a beta version of “Grokipedia,” known by X users as “Grokopedia.”. The world’s richest man said Grokipedia will be published in roughly two weeks, and X News said that the platform will provide “the world’s largest and most accurate knowledge source without centralized control.” – https://cybernews.com/tech/elon-musk-grokipedia-launch/
China develops wearable exoskeleton to train humanoid robots with high accuracy
(DigWatch – 7 October 2025) China’s National University of Defense Technology has partnered with appliance maker Midea Group to introduce HumanoidExo. It is a system with a wearable suit that records full-body motion for robots to learn from humans. The system was introduced in a research paper published last week. Humanoids often lose balance because their training data comes from videos and simulations. HumanoidExo solves that issue by capturing humans’ real-time motion through wearable suits. – https://interestingengineering.com/innovation/chinas-exoskeleton-suit-trains-humanoid-robots
Harvard team builds quantum computer that runs continuously for over two hours
(DigWatch – 7 October 2025) A team of Harvard physicists has built a quantum computing machine that can operate continuously without restarting, achieving a significant milestone in experimental quantum hardware. Until now, quantum computing systems have typically run only for milliseconds or seconds before decoherence or atom loss forces a reset. But in a new setup, the team sustained operation for more than two hours, and they claim that, in theory, it could run indefinitely. – https://dig.watch/updates/harvard-team-builds-quantum-computer-that-runs-continuously-for-over-two-hours – https://www.thecrimson.com/article/2025/10/2/quantum-computing-breakthrough/
AI transforms retinal scans into predictive health tools
(DigWatch – 7 October 2025) AI and machine learning are transforming ophthalmology, making retinal scans powerful tools for predicting health risks. Retinal images offer a non-invasive view of blood vessels and nerve fibres, showing risks for high blood pressure, kidney, heart, and stroke-related issues. With lifestyle-related illnesses on the rise, early detection through eye scans has become increasingly important. – https://dig.watch/updates/ai-transforms-retinal-scans-into-predictive-health-tools – https://www.thenewdaily.com.au/life/health/2025/10/07/ai-eyes-blood-sugar-heart-disease
Energy Robotics Secures $13.5 Million Series A Funding
(AI Insider – 7 October 2025) Energy Robotics raised $13.5M Series A co-led by Blue Bear Capital and Climate Investment to scale its hardware-agnostic AI platform for autonomous robot and drone inspection of critical infrastructure. The company has completed 1M+ inspections across five continents, saving 32,000+ hours of hazardous work for customers including Shell, BP, Repsol, BASF, Merck, and E.ON. Its fleet-orchestration software targets up to 40% OPEX reduction and higher safety via AI analytics for visual/thermal/leak detection, LLM-driven mission control, and secure integration with enterprise systems and digital twins. – https://theaiinsider.tech/2025/10/07/energy-robotics-secures-13-5-million-series-a-funding/
Tech Basel Miami AI Summit To Ignite Next Wave of Innovation
(AI Insider – 7 October 2025) The Tech Basel Miami AI Summit, powered by eMerge Americas and Miami AI Hub, will convene leading figures in artificial intelligence, finance, defense, and healthcare on December 3, 2025, at the Mayfair House Hotel & Garden in Coconut Grove. The by-application-only event will feature keynotes from Clem Delangue of Hugging Face, Cathie Wood of ARK Invest, and Rohit Patel of Meta Superintelligence Labs, alongside panels, workshops, and high-level networking sessions. The summit aims to drive real-world collaboration and innovation across industries while marking the official kickoff for the 2026 eMerge Americas global tech conference. – https://theaiinsider.tech/2025/10/07/tech-basel-miami-ai-summit-to-ignite-next-wave-of-innovation/
CENGN Announces Living Lab Partnership with Morguard and Nokia
(AI Insider – 7 Octrober 2025) CENGN, Morguard, and Nokia launched three Smart Building Living Labs to let Canadian startups test sensor, robotics, and applied AI solutions in real residential, commercial, and retail environments with private 5G and fibre backbones. The initiative is backed by a $45M ISED Strategic Response Fund investment to support 100+ startups/scaleups via Innovation and Adoption Projects with co-investment up to $250K and $500K, respectively. Located at Place Innovation (Montreal), The Bay Club (Toronto), and St. Laurent Shopping Centre (Ottawa), the labs target use cases like energy management, maintenance, safety, and tenant experience to speed commercialization, IP creation, and job growth. – https://theaiinsider.tech/2025/10/07/cengn-announces-living-lab-partnership-with-morguard-and-nokia/
CoreWeave to Acquire Monolith, Expanding AI Cloud Platform into Industrial Innovation
(AI Insider – 7 October 2025) CoreWeave (Nasdaq: CRWV) agreed to acquire Monolith AI to pair its AI cloud with Monolith’s simulation and test-driven ML, creating a full-stack platform to shorten industrial and manufacturing R&D cycles. Monolith’s anomaly detection, test-plan optimization and next-test recommendation tools—used by Nissan, BMW and Honeywell—reduce physical testing and speed engineering decisions, with McKinsey estimating 20–80% R&D efficiency gains. Terms were not disclosed and closing is subject to customary conditions; the deal complements CoreWeave’s OpenPipe and Weights & Biases additions and advances its push to be the default AI cloud for industry (e.g., Aston Martin F1 partnership). – https://theaiinsider.tech/2025/10/07/coreweave-to-acquire-monolith-expanding-ai-cloud-platform-into-industrial-innovation/