Governance and Legislation
When AI Fuels Atrocities — And How It Can Help Prevent Them
(Brittan Heller – Just Security – 31 March 2025) A series of videos began circulating widely across social media platforms in Mali in January 2022, featuring what appeared to be computer-generated voiceovers discussing domestic politics and France’s military presence in the country. As documented by France24, these synthetic media campaigns demonstrated part of an emerging pattern of AI being used to manipulate public opinion in conflict zones. The incident sparked international concern, with media-forensics experts noting how such technology could be deployed to create confusion about military operations and undermine trust in legitimate information sources. Furthermore, these incidents exemplify how the capabilities of generative AI may fundamentally transform conflict dynamics and security landscapes worldwide. To understand these emerging challenges requires first grasping how generative AI differs from traditional artificial intelligence. While earlier AI systems could analyze vast sets of existing content, generative AI creates entirely new material by learning patterns from vast training datasets. Think of it as the difference between a film critic who can analyze movies and a filmmaker who can create entirely new films based on their understanding of cinematic techniques. Modern generative AI systems serve as these digital creators, capable of generating increasingly convincing text, images, audio, and video that can be nearly indistinguishable from human-created content. This ability fuels new and dangerous forms of manipulation, from mass-produced propaganda to hyper-personalized disinformation. Yet, these same capabilities also hold promise for mitigating atrocities, enabling early detection, crisis response, and digital evidence authentication. Understanding both sides of this equation is crucial for crafting effective policy responses. The most sophisticated AI systems today use “transformer architecture” that can understand complex contexts and relationships. These neural networks process information through multiple layers, each refining the output based on patterns learned from training data. The result is synthetic content that becomes increasingly difficult to distinguish from authentic material. –
https://www.justsecurity.org/109612/ai-can-fuel-or-prevent-atrocities/
A Call for US Leadership in the Digital Age
(Center for Democracy & Technology – 31 March 2025) The global digital economy stands at a crossroads. The decisions made today will determine whether the Internet remains an open engine for innovation, economic growth, and free expression, or becomes fragmented and controlled by forces hostile to these values. The United States has the opportunity—and responsibility—to lead the world towards a future where the Internet empowers individuals, businesses, and societies. As a group of organizations and experts that believe an open, global, secure, and trusted Internet is crucial to digital trade and online discourse, we are eager to support the administration in advancing principles that protect the Internet’s ability to enable innovation, promote free expression and access to information, and foster a dynamic digital economy. – https://cdt.org/insights/a-call-for-us-leadership-in-the-digital-age/
Defense, Intelligence, and Warfare
AI for Military Decision-Making. Harnessing the Advantages and Avoiding the Risks
(Emelia Probasco, Helen Toner, Matthew Burtell, Tim G. J. Rudner – Center for Security and Emerging Technology – April 2025) The integration of artificial intelligence into military operations has become a significant focus for armed forces globally. Military commanders are interested in AI’s potential to improve decision-making, especially at the operational level of war, where they must integrate a lot of information quickly to make life-and-death decisions. However, the enthusiasm for AI-enabled decision support systems (DSS) must be balanced with an understanding of their capabilities and limitations to ensure appropriate and effective deployment. This report reviews recently proposed uses of AI-enabled DSS, provides a simplified framework for considering AI-DSS capabilities and limitations, and recommends practical risk mitigations that commanders might employ when operating with an AI-enabled DSS. – https://cset.georgetown.edu/publication/ai-for-military-decision-making/
Army to deploy integrated data layer for NGC2 at division level before next Project Convergence
(Carley Welch – Breaking Defense – 31 March 2025) The Army plans to have its Next Generation Command and Control’s (NGC2) integrated data layer fully operational at the division level before next year’s Project Convergence Capstone (PCC) event, Army senior leaders said. The integrated data layer is a user interface where sensors from multiple domains can work together to supply information on enemy targets and other data the warfighter needs to make important calls. It’s a key part of the technology stack that makes up NGC2, and it creates a framework where, for example, artillery, operational, aviation and other systems can talk to each other so the warfighters operating these systems don’t have to digest the various data separately. – https://breakingdefense.com/2025/03/army-to-deploy-integrated-data-layer-for-ngc2-at-division-level-before-next-project-convergence/
US Marines deploy $5K kamikaze drones with 12-mile range for battlefield strikes
(Kapil Kajal – Interesting Engineering – 31 March 2025) The U.S. Marine Corps took a significant step in modernizing its combat tactics by establishing the Marine Corps Attack Drone Team (MCADT). This initiative, spearheaded by Maj. Gen. Anthony M. Henderson of the Training Command and Brig. Gen. Simon M. Doran of the Marine Corps Warfighting Laboratory aims to address the increasing reliance on armed first-person view (FPV) drone technology observed in recent conflicts, notably in Eastern Europe. The creation of MCADT is seen as a response to the rapid evolution of drone capabilities in warfare. – https://interestingengineering.com/military/us-marines-to-deploy-kamikaze-drones
The Army wants simple, cheap unmanned tech—here are some options
(Meghann Myers – Defense One – 28 March 2025) Russia’s invasion of Ukraine has brought into stark relief that the U.S. Army needs not just powerful vehicles and artillery, but lots and lots of simpler unmanned and counter-unmanned systems for drone warfare. “When America brings its best technology-wise, it’s exquisite, and we should keep doing that,” Gen. Jim Rainey, the head of Army Futures Command, said Thursday at AUSA’s Global Force Symposium. “At the same time, we should buy cheap mass. [There’s a] lot of value in 30mm cheap rounds that can knock down UAVs, even if you have to shoot a burst of 30 at it.”. One company has a 3D-printed unmanned aerial vehicle on offer as part of the Army’s Low Altitude Stalking and Strike Ordnance, or LASSO, program, which is due to put out a request for proposals any day now. – https://www.defenseone.com/technology/2025/03/army-wants-simple-cheap-unmanned-techhere-are-some-options/404141/?oref=d1-featured-river-secondary
Frontiers
OpenAI plans to release a new ‘open’ AI language model in the coming months
(TechCrunch – 31 March 2025) OpenAI says that it intends to release its first “open” language model since GPT‑2 “in the coming months.”. That’s according to a feedback form the company published on its website Monday. The form, which OpenAI is inviting “developers, researchers, and [members of] the broader community” to fill out, includes questions like, “What would you like to see in an open-weight model from OpenAI?” and “What open models have you used in the past?” – https://techcrunch.com/2025/03/31/openai-plans-to-release-a-new-open-language-model-in-the-coming-months/
AI Startup experial Raises €2M to Revolutionize Market Research with Digital Twins
(AI Insider – 31 March 2025) Cologne-based experial, a DeepTech startup transforming market research through AI-powered simulations, has raised €2 million in pre-Seed funding to accelerate development of its digital twin technology. The round was led by Capnamic, with participation from xdeck and angel investors including Professor Bernd Schmitt of Columbia University. – https://theaiinsider.tech/2025/03/31/ai-startup-experial-raises-e2m-to-revolutionize-market-research-with-digital-twins/
Isomorphic Labs Announces $600M Funding to Further Develop its Next-generation AI Drug Design Engine and Advance Therapeutic Programs into the Clinic
(AI Insider – 31 March 2025) Isomorphic Labs raised $600M in its first external funding round led by Thrive Capital, with support from GV and Alphabet, to accelerate its AI-driven drug discovery efforts. Founded by Sir Demis Hassabis, the company uses its next-gen AI drug design engine — including AlphaFold 3 — to develop treatments across various therapeutic areas, with partnerships including Novartis and Eli Lilly. The funding will fuel continued R&D, clinical advancement of internal programs, and global team expansion as Isomorphic aims to redefine drug discovery through AI. – https://theaiinsider.tech/2025/03/31/isomorphic-labs-announces-600m-funding-to-further-develop-its-next-generation-ai-drug-design-engine-and-advance-therapeutic-programs-into-the-clinic/
DeepIP Announces $15M in Funding to Bring Patents to the Age of AI
(AI Insider – 31 March 2025) DeepIP has raised $15M in Series A funding to transform the outdated, manual patent filing process with AI, aiming to improve speed, accuracy, and accessibility for IP professionals. Founded by AI experts with deep legal-tech experience, DeepIP integrates directly into Microsoft Word and helps patent practitioners cut drafting time by up to 50%, while ensuring compliance and data security. Already used to draft 8,500 patents, the platform learns individual writing styles and has seen rapid adoption across top U.S. law firms, positioning itself as the Harvey equivalent for intellectual property. – https://theaiinsider.tech/2025/03/31/deepip-announces-15m-in-funding-to-bring-patents-to-the-age-of-ai/
Zhipu AI Launches Free AI Agent, AutoGLM Rumination, in China’s Intensifying AI Race
(AI Insider – 31 March 2025) Chinese artificial intelligence firm Zhipu AI has launched AutoGLM Rumination, a free AI agent designed to handle complex tasks such as deep research, web searches, travel planning, and report writing. CEO Zhang Peng introduced the product at a Beijing event, highlighting its power by proprietary models GLM-Z1-Air and GLM-4-Air-0414. – https://theaiinsider.tech/2025/03/31/zhipu-ai-launches-free-ai-agent-autoglm-rumination-in-chinas-intensifying-ai-race/
Kay.ai Secures $3M to Eliminate Data Entry Work in Insurance Operations
(AI Insider – 31 March 2025) Kay.ai raised $3M in funding led by Wing VC to scale its AI co-worker, which eliminates manual data entry and operational bottlenecks for insurance brokers and agencies. Its adaptive AI automates complex workflows like submissions and renewals without APIs or major system changes, addressing a $300B BPO market and an impending labor shortage in insurance. Founded by machine learning experts from Microsoft, Twitter, and Amazon, Kay.ai is already delivering major efficiency gains to early partners by saving time, cutting costs, and reducing errors. – https://theaiinsider.tech/2025/03/31/kay-ai-secures-3m-to-eliminate-data-entry-work-in-insurance-operations/
Spark, Mayfield, and Kleiner Perkins pour another $75M into AI data center chipmaker Retym
(TechCrunch – 31 March 2025) Retym, a U.S. chipmaker with Israeli roots, has raised a fresh $75 million series D led by James Kuklinski of Spark Capital. Existing investors — Mayfield’s Navin Chaddha and Kleiner Perkins’ Mamoon Hamid — also participated, bringing its total raised to $180 million, it says. The chipmaker is another startup benefiting from AI’s glow-up. Its chips don’t process AI workloads directly — it’s not a Nvidia GPU competitor. It’s working on a new “programmable coherent digital signal processing DSP” chip that allows the innards of data centers to communicate faster both internally and with external sources. – https://techcrunch.com/2025/03/31/spark-mayfield-and-kleiner-perkins-pour-another-75m-into-ai-data-center-chipmaker-retym/
Taxo Raises $5M from Y Combinator, General Catalyst, and Character to Build Out Autonomous Systems for Healthcare
(AI Insider – 31 March 2025) Taxo, a healthcare AI startup, raised $5M in seed funding led by Y Combinator, General Catalyst, and Character, with participation from top angel investors. Its AI engine automates healthcare administration by reducing paperwork by 90% and achieving 98% decision accuracy, aiming to solve the industry’s massive inefficiency crisis. Founded by physician Dr. Ahmed Kerwan, Taxo is fully HIPAA and SOC2 compliant, with explainable decisions and a mission to bring its technology to hospitals, governments, and large healthcare organizations. – https://theaiinsider.tech/2025/03/31/taxo-raises-5m-from-y-combinator-general-catalyst-and-character-to-build-out-autonomous-systems-for-healthcare/
PhotonDelta and Silicon Catalyst Announce Strategic Collaboration to Drive Innovation for Early-Stage Photonic Startups
(Quantum Insider – 31 March 2025) PhotonDelta and Silicon Catalyst have formed a strategic collaboration to support the global scaling of early-stage photonics startups. The partnership combines PhotonDelta’s $1.5B-backed photonic chip ecosystem with Silicon Catalyst’s semiconductor incubator of over 500 partners. Startups will gain access to technical advisors, go-to-market support, and funding opportunities across both ecosystems, accelerating commercialization in photonics. – https://thequantuminsider.com/2025/03/31/photondelta-and-silicon-catalyst-announce-strategic-collaboration-to-drive-innovation-for-early-stage-photonic-startups/
D-Wave and Japan Tobacco Validate Quantum and AI Workflow Towards Generative Drug Discovery
(Quantum Insider – 31 March 2025) D-Wave and Japan Tobacco’s pharma division completed a proof-of-concept project using a quantum-hybrid workflow to enhance LLMs for generative drug discovery. The integration of D-Wave’s annealing quantum computing with JT’s AI models led to the generation of more valid, “drug-like” molecules compared to classical methods. Molecules produced through QPU-assisted training showed higher drug-likeness scores and better energy efficiency, demonstrating the benefits of quantum computing in generative AI. JT plans to continue developing Quantum AI technologies for drug discovery and expand quantum computing use in molecular design. – https://thequantuminsider.com/2025/03/31/d-wave-and-japan-tobacco-validate-quantum-and-ai-workflow-towards-generative-drug-discovery/
Google Researchers Say Quantum Theory Suggests a Shortcut for Learning Certain Neural Networks
(Quantum Insider – 31 March 2025) A new theoretical study from Google Quantum AI shows that quantum computers could learn certain neural networks exponentially faster than classical algorithms when data follows natural patterns like Gaussian distributions. The researchers developed a quantum algorithm that outperforms classical gradient-based methods in learning “periodic neurons,” a function type common in machine learning. The findings rely on a theoretical framework and require the use of special quantum states, which are not yet practical to prepare on current hardware. – https://thequantuminsider.com/2025/03/31/google-researchers-say-quantum-theory-suggests-a-shortcut-for-learning-certain-neural-networks/
Australian Government Leads QuintessenceLabs’ $20 Million (AUD) Investment Round
(Quantum Insider – 31 March 2025) Australia has invested $15 million AUD into QuintessenceLabs to strengthen its position in quantum cybersecurity and support local manufacturing and global expansion. The funding comes from the National Reconstruction Fund and makes the government the lead investor in a $20 million AUD round. QuintessenceLabs develops quantum-based encryption tools to defend against future cyber threats and serves clients including JPMorgan Chase and U.S. national security agencies. – https://thequantuminsider.com/2025/03/31/australian-government-leads-quintessencelabs-20-million-aud-investment-round/