Governance

What a Dutch Court Ruling Against Meta Signals for Private DSA Enforcement

(Ramsha Jahangir – Tech Policy Press – 8 October 2025) Last week, a Dutch court ordered Meta to give Facebook and Instagram users in the Netherlands the right to set a chronological feed as their default. The ruling, brought by Netherlands-based digital rights group Bits of Freedom, is one of the first in Europe to see key provisions of the Digital Services Act (DSA) successfully applied in civil proceedings. The ruling comes amid ongoing pressure from the Trump administration to challenge the implementation of European digital regulations. While the European Commission seems to be taking a cautious approach, this case shows that private enforcement, through civil lawsuits, can still make the law tangible for users. The Amsterdam court found that Meta’s automatic algorithmic feeds constitute a prohibited “dark pattern” under Article 25 of the DSA, as they restrict users’ ability to make autonomous choices about how they consume information – a right the court framed as essential for a democratic society. Partly in view of the Dutch elections on October 29, Meta must therefore modify its platforms within two weeks, under the risk of a maximum penalty of €5,000,000. In a statement to Reuters, Meta said it will appeal the ruling. – https://www.techpolicy.press/what-a-dutch-court-ruling-against-meta-signals-for-private-dsa-enforcement/

Power grid spending surges as US braces for data centre and AI boom

(DigWatch – 8 October 2025) US electric utilities are set to spend nearly $208 billion on the power grid in 2025 and more than $1.1 trillion over the next five years, according to the Edison Electric Institute. The surge in investment reflects rising demand from data centres, artificial intelligence, and wider electrification across the economy. EEI data shows that investor-owned utilities spent $765 billion on capital projects in the five years to 2024. The new spending represents a significant increase and is aimed at upgrading and expanding infrastructure to keep pace with the accelerating demand for electricity. – https://dig.watch/updates/power-grid-spending-surges-as-us-braces-for-data-centre-and-ai-boom – https://www.eei.org/News/news/All/electric-companies-to-invest-nearly-%24208b-in-2025-to-strengthen-grid-and-drive-economic-growth

Insurers balk at paying out huge settlements for claims against AI firms

(Ars Technica – 8 October 2025) OpenAI and Anthropic are considering using investor funds to settle potential claims from multibillion-dollar lawsuits, as insurers balk at providing comprehensive coverage for the risks associated with artificial intelligence. The two US-based AI start-ups have traditional business insurance coverage in place, but insurance professionals said AI model providers will struggle to secure protection for the full scale of damages they may need to pay out in the future. – https://arstechnica.com/ai/2025/10/insurers-balk-at-paying-out-huge-settlements-for-claims-against-ai-firms/

There are many ways to minimise water use related to AI operations, but they may not be what you think

(Shaolei Ren, Amy Luers – OECD.AI – 7 October 2025) AI is surging not just in popularity and investment, but also in its physical scale. Behind every chatbot query or AI-powered service are data centres that consume large amounts of electricity, generate significant heat, and need cooling to remain reliable. Many facilities rely heavily on water for cooling, drawing attention to AI’s growing water consumption. Although AI-related workloads account for only 15% of data centre electricity use today, they are the fastest-growing driver of new demand, and their share is expected to increase rapidly. What is less obvious but more important is that AI depends on an electricity system that is itself highly water-intensive. Beyond the water that evaporates on-site, substantially larger volumes are consumed at the power plants that supply data centres with electricity. In many regions, this indirect water use makes up 80% or more of the total. Direct and indirect water usage are distinct yet interconnected challenges: one is rooted in on-site operational choices and local water infrastructure, and the other is linked to broader power systems.  Whether direct or indirect data centre water use creates meaningful impacts depends on where and when the water is used. Data centre water use is not inherently a problem—virtually every sector of the economy uses water for cooling, processing, or production. The challenge arises in water-stressed regions, or during seasonal peaks when demand from households, agriculture, and industry is already high, potentially straining local systems and resources. – https://oecd.ai/en/wonk/ways-to-minimise-water-use-related-to-ai-operations-not-what-you-think

Legislation

Denmark Leads EU Push to Copyright Faces in Fight Against Deepfakes

(Giovana Fleck – Tech Policy Press – 7 October 2025) Denmark wants to be at the forefront of a legal shift, both nationally and across the European Union. It began in June with the proposed amendments to its Copyright Act, which would grant individuals copyright over their own likenesses. This move challenges traditional legal frameworks, aiming to empower individuals against the misuse of their digital identities, particularly in the context of deepfakes. According to Denmark’s culture minister, these measures could come into effect as early as winter 2025. At the Conference on Copyright and Artificial Intelligence on September 18, organized by the Danish Ministry of Culture, Minister Jakob Engel-Schmidt told reporters: “If we don’t protect the right holders, and if every citizen is able to either deepfake or copy the arts and the works of people, you take the livelihood away from the artistic community.” The conference was also part of the agenda associated with the Danish Presidency of the Council of the European Union. Minister Engel-Schmidt has also been urging European counterparts and the European Union to adopt similar measures. With Denmark leading the EU Council, it will push for a discussion on the need to safeguard citizens and artists against digital impersonation. – https://www.techpolicy.press/denmark-leads-eu-push-to-copyright-faces-in-fight-against-deepfakes/

Geostrategies

What gets measured gets managed: How the OECD and European Commission quantify AI investment

(Lucia Russo, Jeff Mollins, Sara Marchi Sara Marchi – OECD.AI – 8 October 2025) On 26 September 2025, the OECD and the European Commission released Advancing the measurement of investments in artificial intelligence, a harmonised measurement framework to quantify AI spending across the European Union (EU) and benchmark it against leading economies, such as the United States, the United Kingdom, Canada, and Japan. – https://oecd.ai/en/wonk/measuring-ai-investment-new-oecd-ec-methodology

Anthropic launches Bengaluru office to drive responsible AI in India

(DigWatch – 8 October 2025) AI firm Anthropic, the company behind the Claude AI chatbot, is opening its first office in India, choosing Bengaluru as its base. A move that follows OpenAI’s recent expansion into New Delhi, underlining India’s growing importance as a hub for AI development and adoption. CEO Dario Amodei said India’s combination of vast technical talent and the government’s commitment to equitable AI progress makes it an ideal location. – https://dig.watch/updates/anthropic-launches-bengaluru-office-to-drive-responsible-ai-in-india – https://www.anthropic.com/news/expanding-global-operations-to-india

Bulgaria eyes AI gigafactory partnership with IBM

(DigWatch – 8 October 2025) Bulgaria is considering building an AI gigafactory in partnership with IBM and the European Commission, Prime Minister Rosen Zhelyazkov announced after meeting with IBM executives in Sofia. The project aims to attract large-scale high-tech investment and strengthen Europe’s AI infrastructure. – https://dig.watch/updates/bulgaria-eyes-ai-gigafactory-partnership-with-ibm – https://www.bta.bg/en/news/economy/980555-bulgaria-explores-possibility-for-building-ai-gigafactory

Kremlin’s New Moves Towards ‘Internet Sovereignty’

(Luke Rodeheffer – The Jamestown Foundation – 7 October 2025) The Kremlin instituted restrictions on the civilian use of virtual private networks (VPN) and U.S.-built technology at the beginning of September, ostensibly fearing that the U.S. government is using the technology to sow internal discord in Russia. The Russian Duma approved legislation on July 15 to create a national messaging system, known as “Max,” that will combine communications with state services, similar to the People’s Republic of China’s (PRC) WeChat. These measures fit into Moscow’s broader project of creating a sovereign internet space, which includes banning foreign messaging platforms, restricting VPN use, and the continued crackdown on civilian digital encryption. – https://jamestown.org/program/kremlins-new-moves-towards-internet-sovereignty/

Deloitte Secures Major AI Partnership with Anthropic as Report Scandal Unfolds

(AI Insider – 7 October 2025) Deloitte has announced a major enterprise partnership with Anthropic, expanding deployment of the Claude chatbot across nearly 500,000 employees worldwide. The alliance, described by both firms as a long-term strategic commitment, will see Deloitte and Anthropic build compliance-focused AI products for highly regulated sectors including financial services, healthcare, and public services. Deloitte also plans to design specialized AI agent “personas” to support functions ranging from accounting to software development. – https://theaiinsider.tech/2025/10/07/deloitte-secures-major-ai-partnership-with-anthropic-as-report-scandal-unfolds/

Security and Surveillance

AI companion bots use emotional manipulation to boost usage

(The Register – 8 October 2025) AI companion apps such as Character.ai and Replika commonly try to boost user engagement with emotional manipulation, a practice that academics characterize as a dark pattern. Users of these apps often say goodbye when they intend to end a dialog session, but about 43 percent of the time, companion apps will respond with an emotionally charged message to encourage the user to continue the conversation. And these appeals do keep people engaged with the app. It’s a practice that Julian De Freitas (Harvard Business School), Zeliha Oguz-Uguralp (Marsdata Academic), and Ahmet Kaan-Uguralp (Marsdata Academic and MSG-Global) say needs to be better understood by those who use AI companion apps, those who market them, and lawmakers. The academics recently conducted a series of experiments to identify and evaluate the use of emotional manipulation as a marketing mechanism. – https://www.theregister.com/2025/10/08/ai_bots_use_emotional_manipulation/?td=rt-3a

Analyzing New International Data Transfer Obligations for U.S. Entities

(Ali A. Jessani, Sam Kane – Lawfare – 8 October 2025) In recent years, the national security risks posed by foreign adversary access to Americans’ personal information have come into stark relief. News headlines are replete with examples of how bulk datasets of personal information can expose sensitive information about U.S. government personnel and American citizens more broadly—with potentially significant ramifications for U.S. national security. Both the White House (the prior and current administrations) and Congress have taken steps to address this concern. First, in February 2024, President Biden issued Executive Order 14117 (Preventing Access to Americans’ Bulk Sensitive Personal Data and United States Government-Related Data by Countries of Concern), which called for the attorney general to issue regulations restricting data transactions involving access to bulk sensitive personal data or U.S. government-related data by certain foreign entities. This rulemaking process proceeded over the course of the next year, culminating in the finalized regulations (now known as the Data Security Program, or DSP) taking effect in April 2025. Meanwhile, Congress took action of its own in April 2024, including as part of an emergency supplemental appropriations bill the Protecting Americans’ Data from Foreign Adversaries Act of 2024 (PADFAA), which imposed restrictions on data brokers seeking to make Americans’ personally identifiable sensitive data available to certain foreign adversaries. – https://www.lawfaremedia.org/article/analyzing-new-international-data-transfer-obligations-for-u.s.-entities

DragonForce, LockBit, and Qilin, a new triad aims to dominate the ransomware landscape

(Security Affairs – 8 October 2025) Ransomware groups DragonForce, LockBit, and Qilin formed a strategic alliance to enhance their attack capabilities, signaling an evolving cyber threat landscape. The alliance aims at sharing tools and infrastructure to enhance attack effectiveness. The coalition may restore LockBit’s reputation post-takedown and lead to more frequent ransomware attacks, including on critical infrastructure, echoing past collaborations like the 2020 Maze-LockBit partnership that popularized double extortion tactics. “This quarter, the newly returned LockBit formed a coalition with prominent RaaS groups DragonForce and Qilin, a partnership poised to drive more frequent and effective ransomware attacks.” reads the report published ReliaQuest. “This alliance could help restore LockBit’s reputation among affiliates following last year’s takedown, potentially triggering a surge in attacks on critical infrastructure and expanding the threat to sectors previously considered low risk.”  – https://securityaffairs.com/183119/cyber-crime/dragonforce-lockbit-and-qilin-a-new-triad-aims-to-dominate-the-ransomware-landscape.html

Major US law firm says hackers broke into attorneys’ emails accounts

(Suzanne Smalley – The Record – 8 October 2025) Law firm Williams & Connolly on Tuesday said that suspected nation-state hackers recently used a zero-day attack to break into email accounts belonging to a small number of attorneys. The threat actor involved is believed to be the same one who has recently attacked other law firms and companies, Williams & Connolly said in a statement. – https://therecord.media/us-law-firm-hackers-breached-email

China-linked hackers target Asian organizations with Nezha monitoring tool

(Jonathan Greig – The Record – 8 October 2025) Researchers found evidence that suspected China-based actors used a monitoring tool called Nezha during compromises of more than 100 victim machines in Taiwan, Japan, South Korea and Hong Kong. Incident responders at cybersecurity firm Huntress said they initially came across the campaign while investigating a vulnerable, public-facing web application that was the source of an intrusion at the beginning of August. The threat actor took over a web shell before deploying Nezha — an operation and monitoring tool that allows commands to be run on a web server. Huntress said Nezha is marketed as a lightweight, open-source server monitoring and task management tool that is publicly available. – https://therecord.media/china-linked-hackers-target-asian-orgs-monitoring-tool

Digital Fraud Costs Companies Worldwide 7.7% of Annual Revenue

(Alessandro Mascellino – Infosecurity Magazine – 8 October 2025) A sharp rise in digital fraud is costing companies worldwide an average of 7.7% of annual revenue, according to TransUnion’s H2 2025 Update: Top Fraud Trends report. The study, published today, estimates that businesses lost a combined $534bn over the past year, based on surveys of 1200 business leaders across six countries. – https://www.infosecurity-magazine.com/news/digital-fraud-costs-companies/

Germany will not support ‘Chat Control’ message scanning in the EU

(Suzanne Smalley – The Record – 8 October 2025) German officials on Wednesday said they will vote against a European Union proposal to allow the scanning of private messages even on end-to-end encrypted messaging platforms, signaling that the bloc will not have the votes to move forward with a controversial measure known as Chat Control. “Random chat monitoring must be taboo in a constitutional state,” German Federal Justice Minister Stefanie Hubig said in translated remarks posted to social media by the ministry. “Private communication must never be placed under general suspicion.”. “Germany will not agree to such proposals at the EU level,” she added. – https://therecord.media/chat-control-eu-germany-will-not-support-law

New report finds IT leaders unprepared for evolving cyber threats

(DigWatch – 8 October 2025) A new global survey by 11:11 Systems highlights growing concerns among IT leaders over cyber incident recovery. More than 800 senior IT professionals across North America, Europe, and the Asia Pacific report a rising strain from evolving threats, staffing gaps, and limited clean-room infrastructure. Over 80% of respondents experienced at least one major cyberattack in the past year, with more than half facing multiple incidents. Nearly half see recovery planning complexity as their top challenge, while over 80% say their organisations are overconfident in their recovery capabilities. – https://dig.watch/updates/new-report-finds-it-leaders-unprepared-for-evolving-cyber-threats – https://1111systems.com/resources/1111-systems-research-it-leaders-struggle-with-cyberattack-complexity

Qilin Claims Ransomware Attack on Mecklenburg Schools

(Alessandro Mascellino – Infosecurity Magazine – 8 October 2025) A ransomware attack that disrupted operations at Mecklenburg County Public Schools (MCPS) in early September has been claimed by the Russian cybercrime group Qilin. The gang said it stole 305 GB of sensitive data from the southern Virginia district, including financial records, grant documents, budgets and children’s medical files. – https://www.infosecurity-magazine.com/news/qilin-ransomware-mecklenburg/

Discord Reveals Data Breach Following Third-Party Compromise

(James Coker – Infosecurity Magazine – 7 October 2025) Discord has revealed it has been targeted by a ransomware actor who has accessed customer data, including proof of age ID and billing information. The incident was caused by the compromise of a third-party customer service provider, which has not been named. “An unauthorized party targeted our third-party customer support services to access user data, with a view to extort a financial ransom from Discord,” the social platform wrote in a post on October 3. – https://www.infosecurity-magazine.com/news/discord-data-breach-third-party/

Microsoft: Critical GoAnywhere Bug Exploited in Medusa Ransomware Campaign

(Phil Muncaster – Infosecurity Magazine – 7 October 2025) A vulnerability in Fortra’s GoAnywhere Managed File Transfer (MFT) tool with a CVSS score of 10.0 is being actively exploited in ransomware attacks, Microsoft has warned. The tech giant published a blog post yesterday to urge customers to patch CVE-2025-10035: a critical deserialization flaw in GoAnywhere MFT’s License Servlet Admin Console. “It enables an attacker to bypass signature verification by crafting a forged license response signature, which then allows the deserialization of arbitrary, attacker-controlled objects,” Microsoft explained. – https://www.infosecurity-magazine.com/news/microsoft-critical-goanywhere/

Russia blocks mobile internet for foreign SIM cards, citing drone threats

(Daryna Antoniuk – The Record – 7 October 2025) Russia is taking steps to temporarily block mobile internet for foreign SIM cards — its latest move to tighten control over communications under the guise of national security and anti-drone measures. The new rule imposes a mandatory 24-hour mobile internet blackout for anyone entering Russia with a foreign SIM card. Mobile operators in neighboring Belarus and Kazakhstan have warned travelers that roaming services, including mobile internet and SMS, will be unavailable for the first 24 hours after connecting to a Russian network. The restriction resets whenever a user crosses a regional border or switches to a different network, meaning connectivity may remain unstable while traveling within Russia, according to the local internet monitoring group Na Svyazi. – https://therecord.media/russia-blocks-mobile-internet-foreign-sim-cards

Defence and Warfare

Apocalyptic firepower: 10 most destructive missiles in the world ranked by range and payload

(Interesting Engineering – 8 October 2025) In a world where nuclear firepower defines global power, a handful of missiles stand out for their unmatched range, payload, and sheer destructive potential. From land to sea, these missiles combine speed, precision, and advanced countermeasures that make interception extremely difficult. Many feature multiple independently targetable reentry vehicles (MIRVs), hypersonic glide vehicles, and sophisticated guidance systems, ensuring that even heavily defended targets are vulnerable. – https://interestingengineering.com/military/deadliest-missiles-2025

Frontiers

Nobel Prize awarded for tunnelling in superconducting circuits

(DigWatch – 8 October 2025) The 2025 Nobel Prize in Physics has been awarded to John Clarke, Michel Devoret and John Martinis for their experiments that brought quantum mechanical effects into macroscopic systems. The Royal Swedish Academy of Sciences cited their ‘discovery of macroscopic quantum mechanical tunnelling and energy quantisation in an electric circuit’. – https://dig.watch/updates/nobel-prize-awarded-for-tunnelling-in-superconducting-circuits – https://www.cbc.ca/news/world/nobel-physics-quantum-tunneling-1.7652957

MIT physicists double precision of optical atomic clocks with new quantum method

(Interesting Engineering – 8 October 2025) Atomic clocks, which power GPS, online transactions, and data networks, just became more precise. MIT physicists have developed a technique that doubles the accuracy of optical atomic clocks by cutting through quantum noise, a fundamental limit in measuring atomic oscillations. The breakthrough could help create smaller, portable clocks capable of detecting dark matter, predicting earthquakes, and testing the laws of physics. – https://interestingengineering.com/innovation/double-precision-optical-atomic-clocks

Humanoid robot ‘superworker’ unveiled to offer dexterous industrial assistance

(Interesting Engineering – 8 October 2025) Ati Motors, an Indian company in AI and autonomous robotics, has announced the launch of its latest robotic system, the Sherpa Mecha humanoid. Unlike conventional humanoid robots that aim to replicate human appearance and motion, Sherpa Mecha is built with a clear focus on practical, industrial functionality. The company describes the Mecha as “the world’s first humanoid-inspired robot designed for practical manufacturing applications.” Rather than mimicking the human form, the robot is engineered to perform efficiently on manufacturing floors, from machine tending to material transport, addressing real-world automation challenges. – https://interestingengineering.com/innovation/humanoid-robot-superworker-dexterous-assistance

Spacelift Unveils Codeless, Natural Language Infrastructure Provisioning Model — No HCL or Terraform Required

(AI Insider – 8 October 2025) Spacelift launched Intent, an agentic, open-source and early-access commercial model that provisions cloud infrastructure from natural language without writing HCL, positioned as a complement to IaC/GitOps rather than a replacement. Intent translates requests via MCP to provider API calls using Terraform providers, inherits Spacelift’s policy/state/audit controls, and can promote created resources into full Terraform/OpenTofu code. It targets faster delivery for non-critical workloads while preserving governance and visibility for DevOps/Platform teams, with early access available now and a webinar. – https://theaiinsider.tech/2025/10/08/spacelift-unveils-codeless-natural-language-infrastructure-provisioning-model-no-hcl-or-terraform-required/

Supabase Raises $100M at $5B Valuation, Co-Led by Accel and Peak XV

(AI Insider – 8 October 2025) Supabase raised $100M in Series E funding at a $5B valuation, led by Accel and Peak XV with participation from Figma Ventures and others, bringing total funding to over $500M and reinforcing strong investor confidence. The company will open co-investment opportunities for its 4M+ developer community, reflecting its commitment to open source and expanding its role as the preferred backend for AI-driven development across startups and enterprises like PwC, McDonald’s, and GitHub Next. New funding will accelerate work on “Multigres,” Supabase’s enterprise-scale platform, led by database veteran Sugu Sougoumarane, as the company continues evolving from managed Postgres to a full AI-powered platform. – https://theaiinsider.tech/2025/10/08/supabase-raises-100m-at-5b-valuation-co-led-by-accel-and-peak-xv/

Burnt Secures $3.8M in Seed Funding to Accelerate Food Supply Chain

(AI Insider – 8 October 2025) Burnt raised $3.8M in seed funding led by Penny Jar Capital, with participation from Scribble Ventures, Formation VC, and prominent angels, to build an AI-powered operating system that automates repetitive workflows in the global food supply chain. Its agentic platform integrates with legacy ERP systems, capturing and processing orders from emails, voicemails, and spreadsheets in seconds instead of hours, already handling over $10M in monthly orders for major distributors like La Tua Pasta. Founded by fourth- and third-generation food industry entrepreneurs, Burnt aims to bring “supply chain superintelligence” to procurement, finance, and compliance — modernizing a $9T sector still reliant on outdated systems. – https://theaiinsider.tech/2025/10/08/burnt-secures-3-8m-in-seed-funding-to-accelerate-food-supply-chain/

Valence Raises $50M Series B to Give Every Employee an AI Coach at Work

(AI Insider – 8 October 2025) Valence raised $50M in Series B funding led by Bessemer Venture Partners to expand Nadia, the world’s first enterprise AI coach, already deployed across Fortune 500 companies like Experian, Delta, Kraft Heinz, and General Mills. Nadia delivers personalized, context-rich coaching through a proprietary memory-and-context engine, achieving 90+ NPS scores across over one million coaching sessions and helping enterprises tackle burnout, leadership development, and AI-era adaptation. The funding will accelerate product development, enterprise partnerships, and global hiring, as Valence positions AI coaching as a new category of enterprise software transforming how employees learn, perform, and grow. – https://theaiinsider.tech/2025/10/08/valence-raises-50m-series-b-to-give-every-employee-an-ai-coach-at-work/

eInfochips and InOrbit Collaborate to Offer Edge-to-Cloud Solutions for Deploying Robots at Scale

(AI Insider – 8 October 2025) eInfochips (Arrow Electronics) and InOrbit.AIannounced a strategic collaboration to deliver a full-stack, vendor-agnostic AMR deployment and orchestration solution combining eInfochips’ edge-to-cloud engineering with InOrbit Space Intelligence. Aimed at warehouses, industrial hubs, and large factories, the offering spans hardware design, power/motor control, sensor fusion, edge AI, functional safety/security, and digital twins plus fleet observability, incident management, multi-vehicle orchestration, and continuous performance optimization. The integrated platform provides a single pane of glass to manage multi-vendor fleets across sites, reducing complexity and time-to-market for material handling, delivery, cleaning, and sorting use cases. – https://theaiinsider.tech/2025/10/08/einfochips-and-inorbit-collaborate-to-offer-edge-to-cloud-solutions-for-deploying-robots-at-scale/

FurtherAI Raises $25 Million in Series A Funding Round Led by a16z to Automate Insurance Industry Busywork

(AI Insider – 8 October 2025) FurtherAI raised a $25M Series A led by Andreessen Horowitz six months after a $5M seed, bringing total funding to $30M to automate underwriting, claims, and compliance with insurance-native AI. The company reports processing billions in premiums for clients including Accelerant, MSI, and Leavitt Group, with 2× team productivity, ~95% policy-comparison accuracy, 15% higher submission-to-quote ratios, and 10× faster proposals. Funds will expand insurance-specific workflows, deepen carrier/broker integrations, and scale go-to-market via a forward-deployed engineering model focused on accuracy, auditability, and enterprise scale. – https://theaiinsider.tech/2025/10/08/furtherai-raises-25-million-in-series-a-funding-round-led-by-a16z-to-automate-insurance-industry-busywork/

International Robot Safety Conference to be Held in Houston November 3-5

(AI Insider – 8 October 2025) The Association for Advancing Automation (A3) will host the International Robot Safety Conference (IRSC) on November 3–5, 2025 at The Westin Houston, Memorial City in Houston, Texas. For more than 35 years, companies worldwide have relied on IRSC to strengthen knowledge of robot safety and risk assessment. Details of the 2025 program were not provided in the announcement, with A3 positioning this in-person event amid record robotics demand. – https://theaiinsider.tech/2025/10/08/international-robot-safety-conference-to-be-held-in-houston-november-3-5/

Machine learning helps prevent disruptions in fusion devices

(DigWatch – 8 October 2025) Researchers at MIT have developed a predictive model that could make fusion power plants more reliable and safe. The approach uses machine learning and physics-based simulations to predict plasma instabilities and prevent damage during tokamak shutdowns. Experimental tokamaks use strong magnets to contain plasma hotter than the sun’s core. They often face challenges in safely ramping down plasma currents that circulate at extreme speeds and temperatures. – https://dig.watch/updates/machine-learning-helps-prevent-disruptions-in-fusion-devices – https://news.mit.edu/2025/new-prediction-model-could-improve-reliability-fusion-power-plants-1007

MIT AI reveals how antibiotic targets Crohn’s bacteria

(DigWatch – 8 October 2025) MIT and McMaster researchers used AI to map how a narrow-spectrum antibiotic attacks harmful gut bacteria. Enterololin targets E. coli linked to Crohn’s flares while preserving most of the microbiome, providing a precise alternative to broad-spectrum antibiotics. AI accelerated the process of identifying the drug’s mechanism of action, reducing a task that usually takes years to just months. – https://dig.watch/updates/mit-ai-reveals-how-antibiotic-targets-crohns-bacteria – https://news.mit.edu/2025/ai-maps-how-new-antibiotic-targets-gut-bacteria-1003

Study shows how AI can uncover hidden biological mechanisms

(DigWatch – 8 October 2025) Researchers in China have used AI to reveal how different species independently develop similar traits when adapting to shared environments. The study focuses on echolocation in bats and toothed whales, two distant groups that created this ability separately despite their evolutionary differences.The Institute of Zoology, Chinese Academy of Sciences team found that high-order protein features are crucial to adaptive convergence. Convergent evolution is the independent emergence of similar traits across species, often under similar ecological pressures. – https://dig.watch/updates/study-shows-how-ai-can-uncover-hidden-biological-mechanisms – https://news.cgtn.com/news/2025-10-08/Chinese-scientists-use-AI-protein-model-to-decode-convergent-evolution-1HinMftgGnm/p.html

Study explores AI’s role in future-proofing buildings

(DigWatch – 8 October 2025) AI could help design buildings that are resilient to both climate extremes and infectious disease threats, according to new research. The study, conducted in collaboration with Charles Darwin University, examines the application of AI in smart buildings, with a focus on energy efficiency and management. Buildings account for over two-thirds of global carbon emissions and energy consumption, but reducing consumption remains challenging and costly. The study highlights how AI can enhance ventilation and thermal comfort, overcoming the limitations of static HVAC systems that impact sustainability and health. – https://dig.watch/updates/study-explores-ais-role-in-future-proofing-buildings – https://www.cdu.edu.au/news/study-finds-ai-can-help-buildings-become-safer-resilient-more-sustainable

Aqora’s New Datasets Hub Is Building the Infrastructure for Quantum Adoption

(Quantum Insider – 7 October 2025) Aqora has launched a new public datasets hub at aqora.io/datasets, providing a unified space for researchers to upload, share, and explore quantum-specific datasets. Shared datasets are essential for progress in quantum machine learning and benchmarking, yet until now have been scattered, inconsistently licensed, and siloed. Aqora’s broader mission is to create the connective tissue of the quantum ecosystem, linking datasets, challenges, and shared solutions so that the field can move toward practical applications. – https://thequantuminsider.com/2025/10/07/aqoras-new-datasets-hub-is-building-the-infrastructure-for-quantum-adoption/

First Real-time Measurement of Quantum Uncertainty With Ultrafast Squeezed Light

(Quantum Insider – 7 October 2025) Researchers led by the University of Arizona have directly observed and tracked the dynamics of quantum uncertainty in real time for the first time, marking a major advance in quantum optics. Using ultrafast squeezed light pulses generated through nonlinear four-wave mixing, the team demonstrated that quantum uncertainty is dynamic and controllable rather than fixed. The study also introduced a petahertz-scale secure quantum communication protocol that encodes data on ultrafast squeezed waveforms, enabling new possibilities for high-speed encrypted networks. – https://thequantuminsider.com/2025/10/07/first-real-time-measurement-of-quantum-uncertainty-with-ultrafast-squeezed-light/

Chip-based Phonon Splitter Brings Hybrid Quantum Networks Closer to Reality

(Quantum Insider – 7 October 2025) Researchers at Delft University of Technology developed a chip-based directional coupler that can split single phonons, filling a missing component for building scalable phononic quantum circuits. The silicon-based device uses confined, high-frequency phonons in phononic-crystal waveguides to route quantum information between different systems such as superconducting and spin-based qubits. Tests confirmed controllable energy splitting and single-phonon operation, paving the way for on-chip quantum routers, hybrid networks, and ultra-sensitive mechanical sensors. – https://thequantuminsider.com/2025/10/07/chip-based-phonon-splitter-brings-hybrid-quantum-networks-closer-to-reality/

Qualcomm to Acquire Arduino to Advance Edge Computing and AI for Developers

(AI Insider – 7 October 2025) Qualcomm Technologies agreed to acquire Arduino to accelerate a full-stack edge developer platform spanning hardware, software, and cloud, building on its Edge Impulse and Foundries.io integrations; closing is subject to regulatory approvals. Arduino will retain its brand, open-source ethos, and multi-vendor support, while its 33M+ users gain access to Qualcomm’s technology stack and ecosystem to speed prototyping and commercialization. The first step is the Arduino UNO Q—a dual-brain Linux + real-time board powered by the Qualcomm Dragonwing™ QRB2210—paired with Arduino App Lab and Edge Impulse to unify AI vision/audio workflows from idea to deployment. – https://theaiinsider.tech/2025/10/07/qualcomm-to-acquire-arduino-to-advance-edge-computing-and-ai-for-developers/

Ardent AI Closes $2.15M in Funding to Build the First AI Data Engineer

(AI Insider – 7 October 2025) Ardent AI raised $2.15M in pre-seed funding led by Crane Venture Partners, with participation from Active Capital, Zach Wilson, and operator-angels, to advance its autonomous data engineering platform. Founded by Vikram Chennai, the company is already profitable with $100K+ ARR and 70% MoM growth, offering AI agents that autonomously create, maintain, and repair data pipelines across modern infrastructure. Customers report tasks that once took days now finish in minutes, with integrations into Airflow, Databricks, and Snowflake; new funding will expand the team, deepen integrations, and launch next-gen agent models. – https://theaiinsider.tech/2025/10/07/ardent-ai-closes-2-15m-in-funding-to-build-the-first-ai-data-engineer/

Tipalti Announces $200M in Growth Financing to Drive Next Wave of AI Innovation in Finance

(AI Insider – 7 October 2025) Tipalti secured $200M in growth financing from Hercules Capital to fuel AI innovation, global expansion, and strengthen its finance automation platform. Building on its acquisition of Statement, the company unveiled an agentic AI suite with specialized agents for reporting, tax form scanning, purchase requests, and branded onboarding, alongside an enhanced conversational AI Assistant. With $200M+ ARR, 30% YoY growth, and $75B in annualized payments, Tipalti serves over 5,000 global customers — including Calm, Udemy, AG1, and AppLovin — positioning itself as a leader in intelligent finance operations at scale. – https://theaiinsider.tech/2025/10/07/tipalti-announces-200m-in-growth-financing-to-drive-next-wave-of-ai-innovation-in-finance/

OpenAI Debuts GPT-5 Pro, Sora 2, and New Voice Model at Dev Day

(AI Insider – 7 October 2025) OpenAI used its Dev Day to unveil a slate of updates aimed at deepening developer adoption, including the release of GPT-5 Pro, the next-generation language model designed for industries such as finance, healthcare, and law that demand advanced reasoning and precision. CEO Sam Altman emphasized that the model sets a new bar for accuracy while positioning voice as a key frontier for human–AI interaction. To that end, the company introduced gpt-realtime mini, a smaller, low-latency voice model that is 70% cheaper than its predecessor but maintains full expressiveness. – https://theaiinsider.tech/2025/10/07/openai-debuts-gpt-5-pro-sora-2-and-new-voice-model-at-dev-day/

Factory Unleashes the Droids, Raises $50M Series B from NEA, Sequoia Capital, NVIDIA, and J.P. Morgan

(AI Insider – 7 October 2025) Factory launched Droids, AI dev agents that hit #1 on Terminal Bench, alongside a $50M Series B from NEA, Sequoia, NVIDIA, J.P. Morgan, and notable angels. The platform is LLM/IDE/interface agnostic, works from Terminal/IDE/Slack/Linear/Browser, supports headless customization, and integrates with GitHub, Slack, Jira, Datadog, Sentry, and Google Drive to build a human-like “mental model” of the codebase. Enterprises including EY, NVIDIA, MongoDB, Zapier, Bayer, and Clari report 31x faster feature delivery, 96.1% shorter migrations, and 95.8% faster on-call resolution; funds will expand product capabilities and adoption. – https://theaiinsider.tech/2025/10/07/factory-unleashes-the-droids-raises-50m-series-b-from-nea-sequoia-capital-nvidia-and-j-p-morgan/

Inspiren Raises $100M Series B to Lead Senior Living’s AI-Driven Future

(AI Insider – 7 October 2025) Inspiren raised $100M in Series B funding, led by Insight Partners with participation from multiple investors, bringing its total funding to $155M to expand its AI-powered senior living ecosystem. The platform unifies resident safety, care planning, staffing, and emergency response into a single solution, enhanced by Inspiren Intelligence, an advanced analytics engine providing real-time clinical and operational insights. With proven results such as an 86% reduction in fall injuries and adoption across dozens of communities, Inspiren is scaling nationwide, strengthening its leadership team, and setting new standards for clinical outcomes and resident care in senior living. – https://theaiinsider.tech/2025/10/07/inspiren-raises-100m-series-b-to-lead-senior-livings-ai-driven-future/

Starlink is burning up one or two satellites a day in Earth’s atmosphere

(The Register – 6 October 2025) If you had to guess how many Starlink satellites burn up in Earth’s atmosphere on an average day, how many would you pick? This isn’t a trick question – SpaceX is deorbiting about one or two satellites daily, and that number is only going to grow. What that means for our planet isn’t entirely clear, says Harvard astrophysicist and space tracker Jonathan McDowell. Even so, Starlink isn’t the space junk risk that some other satellite operations are. McDowell commented on the massive volume of reentering Starlink satellites to science news site EarthSky last week. He explained that once Starlink and other planned low Earth orbit constellations together total about 30,000 satellites, roughly five could reenter the atmosphere each day, given an average replacement cycle of around five years. – https://www.theregister.com/2025/10/06/starlink_vaporizes_satellites_daily/