Daily Digest on AI and Emerging Technologies (15 September 2026) – https://pam.int/daily-digest-on-ai-and-emerging-technologies-15-september-2026/

Daily Digest on AI and Emerging Technologies (16 September 2026) – https://pam.int/daily-digest-on-ai-and-emerging-technologies-16-september-2026/

Daily Digest on AI and Emerging Technologies (17 September 2026) – https://pam.int/daily-digest-on-ai-and-emerging-technologies-17-september-2026/

Daily Digest on AI and Emerging Technologies (18 September 2026) – https://pam.int/daily-digest-on-ai-and-emerging-technologies-18-september-2026/

 

Goyernance, Regulation, Legislation, Geostrategies

Parents, privacy, parliaments: The battle begins for von der Leyen’s social media ban – Ursula von der Leyen’s bill to keep young Europeans off social media is an ambitious response to the growing outcry over the potential harm of TikTok, Instagram and other online platforms. Making the restrictions a reality will require a whole new level of ambition. The European Commission must now negotiate with national governments and technologists to figure out exactly how to impose a law across Europe that would ban kids under 13 from social media, and severely restrict access for those under 15 — all without trampling on privacy of young and old alike in the process of verifying users’ age. Reducing the harms of social media on youth has been a top priority for the Commission president, a medical doctor and mother of seven who has repeatedly decried the internet’s impact on the health of minors since her 2025 State of the Union speech. “If a platform wants to access the children or the teenagers, the minors, they have to prove first that they are safe,” she told reporters on Thursday as she officially unveiled the EU Kids Act, the day after again touting her leadership on the issue in this year’s annual address. The EU’s proposed measures come after Australia imposed the first-ever national social media ban in December 2025, blocking under-16s from creating accounts on major platforms such as Instagram and YouTube. Australia’s ban has had limited impact so far, with a large percentage of kids still having access to social media, which Canberra blames on the platforms’ lack of compliance.  A fierce global campaign by parents and mental health professionals has been warning that social media platforms lead to a range of ills including addiction, social isolation and depression. The “Lost Screen Memorial”, an installation of giant illuminated smartphones representing 50 children who tragically lost their lives following harm linked to social media, displayed at Place des Nations, ahead of the 79th World Health Assembly in Geneva, on May 17, 2026. Child rights campaigners and some EU governments have argued that blanket bans are not the way to go, violating young people’s rights to information and expression. Their concerns were integrated into the new proposal, which attempts to reconcile sometimes conflicting considerations and targets online channels beyond traditional social media, including AI chatbots and video games. But the nuances in the legislation will only add to the challenge of making the law a reality. The EU’s proposal would effectively bar anyone under 13 from creating accounts on social media and video-sharing platforms, while severely restricting teens aged 13 to 15 from using these services independently. Services like video game platforms and AI chatbots have to follow safety-by-design principles that don’t encourage compulsive use among kids. – https://www.politico.eu/article/parents-privacy-parliaments-the-battle-begins-on-von-der-leyens-social-media-ban/

AI companies worry about the tech harming humans in the future. But it may damage the US economy even sooner – President Donald Trump is suddenly at odds with AI’s leading executives over the largely unchecked development of AI technology. And it may be because US economy can’t afford a slowdown. It might seem politically ill-advised, less than two months before the midterm elections to so forcefully stake out a position that’s out of step with the majority of Americans, who overwhelmingly oppose data centers being built near them and remain suspicious of AI’s use cases. But Trump’s pedal-to-the-metal approach suggests he may be more concerned about a bigger question: What happens to the US economy if the AI fever breaks on his watch? There are a lot of ways to slice the metrics on just how dependent the US economy has become on AI (and AI-adjacent) spending. ING estimates tech investments dominated by AI and data centers account for a third of year-over-year economic growth in 2026. Goldman Sachs’ chief equity strategist recently told CNBC that AI investment is driving half of all the profit growth in the S&P 500. – https://edition.cnn.com/2026/09/18/economy/trump-ai-economy

Inside the scramble for trusted AI cops  – A host of industry players, policy wonks and businesses are grappling with the question of how to regulate AI in a trustworthy manner. Inside the White House, it’s mostly business as usual. The industry’s mad dash is the product of an ad hoc regulatory apparatus and a general consensus that swift action from Washington is unlikely in the near term. AI CEOs have been grappling with how to test advanced AI and work with Beijing on safety standards, while the White House favors a solution where the industry finds ways to police itself. – https://www.axios.com/2026/09/18/ai-safety-evaluators-metr-white-house-trump

A global AI safety strategy depends on US-China cooperation. They each see the other as the problem – As concern rises over the risks of artificial intelligence, hopes for any kind of global approach depend on cooperation between the U.S. and China — superpowers that seem only to be more skeptical of each other’s AI strategy. AI governance is expected to be on the agenda in a meeting planned for next week in Washington between President Donald Trump and Chinese leader Xi Jinping. The topic has gained urgency in both countries, with American tech leaders saying that averting worst-case scenarios for rogue AI depends on a degree of cooperation with China. But both countries are vying for a decisive advantage on AI, with Trump warning that any efforts to regulate the technology would help China. That sets up a meeting in which any agreements might be superficial but a possible start to something larger. “A formal agreement or consensus may be near impossible, but they do not need to go that far,” said Samm Sacks, a senior fellow at the Johns Hopkins School of Advanced International Studies’ Institute for America, China, and the Future of Global Affairs. “Trump and Xi just need to create political space by acknowledging AI poses risks to both countries”. – https://apnews.com/article/china-us-ai-safety-trump-xi-cd680bd072ec95e5f4b9c99dbee91319

China is building an AI system for tailoring its foreign propaganda – China’s propaganda is often mocked as blunt, loud and tone-deaf. But state-backed researchers are trying to make it more sophisticated. They are building AI systems that model the cultural and political values of foreign audiences, simulate how they might respond to events and messages and use those predicted reactions to refine what China says to the world. Among the clearest examples is a system unveiled in April 2026 called Huanyu Zhiying 1.0 寰舆智应 – a name that puns ‘the whole world’ with ‘global public opinion’ –at Sun Yat-sen University in Guangzhou. Built from more than 10,000 simulated user profiles spanning a claimed 18 countries, it attempts to model how foreign audiences and media might respond to events and messages. But the system does not merely try to predict what foreigners might think. It is being designed to help formulate what China should say in response. Responses from real Australians helped build its synthetic population. One source of data is Sun Yat-sen University’s survey The World Looks at China, which polled respondents in eight countries – Australia, the United States, Britain, Canada, India, Malaysia, France and South Africa – using samples designed to reflect each country’s demographic composition. The developers say they combined those results with data from the World Values Survey – a long-running international survey of social and political attitudes – and public social-media data to construct Huanyu Zhiying’s simulated foreign users. –  https://www.aspistrategist.org.au/china-is-building-an-ai-system-for-tailoring-its-foreign-propaganda/

Emerald AI, Google, and NVIDIA launch AI Energy Management Alliance – Emerald AI, Google, and NVIDIA announced the launch of the AI Energy Management Alliance (AEMA), a coalition intended to advance data centres that can dynamically manage their electricity use in response to grid conditions. The alliance describes its aim as building AI infrastructure that ‘doesn’t just connect to the grid but works with it’. AEMA describes its approach as technology-neutral and performance-based, focusing on measurable factors such as response speed, duration, predictability, and behaviour during emergencies rather than prescribing specific technologies. Its principles include setting clear requirements for how data centres should operate during grid disturbances, reduce electricity demand when requested and respond during emergencies; standardising technical requirements, performance metrics and operational data sharing; creating faster, risk-adjusted connection pathways for customers that make credible and verifiable flexibility commitments; and allocating interconnection costs according to a facility’s actual impact on the power system. – https://dig.watch/updates/ai-energy-management-alliance-launch

India questions Meta’s intermediary status under the IT Act – India’s Ministry of Electronics and Information Technology (MeitY) is reportedly challenging Meta’s intermediary classification under the IT Act. Officials argue that commercial ad monetisation and algorithmic models and amplification compromise platform neutrality, removing statutory safe-harbour protections. The debate follows domestic investigations into child safety and the commercial promotion of illicit material online. Senior ministry officials announced that Meta, which operates Instagram and Facebook, should be categorised as an active service provider. The challenge comes after Meta agreed to report child safety reports directly to the Indian Cybercrime Coordination Centre, departing from its previous protocol of transmitting notifications exclusively via the US-based National Center for Missing and Exploited Children (NCMEC). – https://dig.watch/updates/indias-questions-meta-status-under-the-it-act

UNESCO explores scaling AI learning across African higher education – UNESCO and its partners are examining how artificial intelligence and simulation-based learning can move beyond pilot projects and become sustainable parts of higher education across Africa, particularly in institutions facing limited connectivity and infrastructure. The issue was discussed during a Strategy Lab organised by the UNESCO Regional Office for Southern Africa and Nudle during UNESCO Digital Learning Week 2026 in Paris. Participants considered how emerging technologies could be integrated into institutions facing constraints including unreliable internet access, limited specialised training equipment, restricted budgets, and uneven educator readiness. – https://dig.watch/updates/unesco-ai-learning-african-education

Queensland’s Fraser Coast council backs data centre water and power safeguards – On 16 September 2026, the Fraser Coast Council in Queensland, Australia unanimously supported a motion concerning the possible development of data centres in the region. The motion was introduced by Councillor Paul Truscott. It recognises growing demand for data centre facilities and calls for measures to prevent adverse impacts on local water and electricity supplies. Under the approach set out in the motion, any prospective data centre would be expected to arrange its own electricity supply and to avoid reliance on the Fraser Coast drinking water network. Councillor Truscott identified recycled or non‑potable water as possible alternatives for data centre operations. – https://dig.watch/updates/fraser-coast-data-centre-water-power-safety

Crime, Terrorism, Security, Surveillance

Why AI Fluency Is Now an Imperative in Closing the Cyber Skills Gap – Artificial intelligence is fundamentally reshaping the cyber threat landscape. Cybercriminals increasingly have access to the same enterprise-grade AI technologies organizations use to defend their digital enterprises. The weaponization of AI is redefining how attacks are designed and executed autonomously, intensifying the pressure on already stretched security teams. As Frontier AI models continue to evolve, security leaders will have less time to modernize their cyber defenses while attackers discover and exploit vulnerabilities across a fast-expanding attack surface at unprecedented speed and scale. Across industries, security leaders’ ability to secure increasingly complex AI and hybrid environments, defend against a rapidly evolving threat landscape and manage mounting operational pressures is being impacted by overwhelmed teams and a shortage of skilled cybersecurity talent. This is weakening their organization’s overall security posture. To maintain stakeholder trust, organizations must keep pace with attackers by shifting from reactive, siloed, human-paced security functions toward a proactive, integrated a nd machine-speed security spanning prevention, response and recovery. Achieving this requires unified, enterprise-wide visibility of cyber risk and a simplified, cohesive security fabric across people, processes and technology. By leveraging AI and autonomous agents within the right governance framework, security leaders can enhance detection and response speed, improve accuracy and reduce the burden of routine tasks on security analysts. However, realizing this potential requires more than deploying new technologies. It requires building AI fluency across the cybersecurity workforce, equipping teams to understand how AI systems operate, where they introduce new risks, how attackers exploit them and how defenders use them responsibly to strengthen security outcomes. In this environment, AI fluency is becoming as essential to cybersecurity professionals as security expertise itself. The cybersecurity skills gap is no longer just a workforce challenge; it is becoming an AI readiness challenge. – https://www.infosecurity-magazine.com/opinions/ai-fluency-cyber-skills-gap/

CISA Upgrades Vulnerability Reporting Platform with More Automation – The US Cybersecurity and Infrastructure Security Agency (CISA) has upgraded its vulnerability reporting and coordination platform to allow for more automation and streamlined processes, as well as new built-in tools that vulnerability researchers can use. Since 2020, CISA has been using Carnegie Mellon University’s Vulnerability Information and Coordination Environment (VINCE), a vulnerability management platform developed that same year by the Computer Emergency and Response Team Coordination Center (CERT/CC), a unit of the university’s Software Engineering Institute (SEI). From September 17, 2026, the US cybersecurity agency is now using VINCE – New Technology (VINCE-NT). “VINCE-NT is a modernized, CISA-managed platform for vulnerability reporting and coordination. It improves how vulnerability reporters, product suppliers and CISA case managers collaborate throughout the disclosure process,” CISA said in an announcement published on social media on September 17. The agency also said the change shifts ownership, sponsorship and management of the platform to its Coordinated Vulnerability Disclosure (CVD) team and enables improved integration with its internal tools and processes. – https://www.infosecurity-magazine.com/news/cisa-upgrades-vulnerability/

Manufacturing Accounts for 22% of all Ransomware Victims – Manufacturing organizations made up over a fifth (22%) of all ransomware victims in the period from April 2025 to March 2026, according to new Black Kite study. This made the industry the most targeted by ransomware attacks for the fifth consecutive year, the analysis found. The volume of ransomware incidents in the manufacturing sector also surged by around 40% year-over-year in the period from January 1 to July 29, 2026 – from 847 to 1183. This growth continued a trend observed by the researchers of disclosed ransomware incidents in manufacturing rising every year since 2022. Manufacturing is viewed as a high-value target for ransomware actors for a number of reasons. One factor is that a successful attack can result in significant downtime for victim organizations, which in turn leads to substantial financial losses, potentially making them more likely to agree to pay an extortion demand to resume operations. The financial impact of such incidents was highlighted in the attack on UK car manufacturing giant Jaguar Land Rover (JLR) in 2025, which is estimated to have cost the UK economy £1.9bn ($2.5bn). Another factor is growing IT-OT convergence in the sector, making it easier for threat actors to compromise industrial systems. – https://www.infosecurity-magazine.com/news/manufacturing-22-ransomware-victims/

New Settra Ransomware Variant Deployed in Attacks on Retail and Manufacturing – A new ransomware variant named Settra has been deployed in incidents targeting the retail and manufacturing sectors, according to Huntress. The variant was first observed in June, and Huntress researchers highlighted notable post-compromise techniques used by threat actors deploying Settra in attacks against an organization in the consumer services and retail sector in July, and a manufacturing firm in September. These techniques include deploying remote monitoring and management (RMM) tools for persistent access, efforts to disable the victims’ recovery options, and installing bring your own vulnerable driver (BYOVD) to impacted systems. The Huntress blog, published on September 17, highlighted how previous research on Settra showed the variant was used for double-extortion tactics, with the attackers threatening to release sensitive corporate information alongside encrypting systems. There is currently not enough evidence to state that Settra is a ransomware-as-a-service (RaaS) operation, the researchers noted. – https://www.infosecurity-magazine.com/news/settra-ransomware-retail/

Hacking group ‘NightEagle’ targeting China’s high-tech sector expands operations to Russia – A cyberespionage group previously known for targeting sensitive technology and defense organizations in China has expanded its operations to Russian companies, according to new research released this week. The group, known as NightEagle or APT-Q-95, has been active since at least 2023 but had previously focused its attacks in Asia. Over the past year, Russian cybersecurity firm Kaspersky said it investigated several incidents involving the group at Russian businesses. In most cases, the hackers used stolen credentials to gain access to corporate networks through virtual private networks, or VPNs. Once inside a network, NightEagle targeted Microsoft Exchange email servers and installed a backdoor known as GhostContainer, which allows attackers to remotely control compromised servers, evade some Windows security and logging mechanisms and redirect network traffic. – https://therecord.media/hacking-group-nighteagle-expands-russia-china

Nations take action on North Korean IT workers after UN report – Multiple countries have taken legal action against North Koreans or local handlers following a report from the United Nations about Pyongyang’s illicit IT worker scheme. The Multilateral Sanctions Monitoring Team (MSMT) — a U.S.-led international committee tasked with tracking compliance of UN sanctions on the Democratic People’s Republic of Korea (DPRK) — released a new report on Wednesday spotlighting the thousands of North Korean nationals who work outside of the country in various industries. The study is an expansion of a 140-page report released last October by the UN that focused specifically on the IT worker scheme — where North Korean nationals steal or purchase IDs to get hired in high-paying IT roles typically at companies. The UN said teams of the IT workers live illegally in China as well as about 40 other countries. – https://therecord.media/nations-take-action-on-north-korean-it-worker-schemes

North Korean hackers infect thousands of devices across 100 countries as part of ‘WaterPlum’ campaign – More than $10.5 million has been stolen by North Korean hackers targeting job seekers as part of a long-running cyber campaign to infiltrate tech companies and fill Pyongyang’s coffers with illicitly gained funds. The FBI and Defense Department partnered with Japan’s National Police Agency and law enforcement agencies in Australia and Germany on a new advisory about “WaterPlum” — a group of cyber actors allegedly stealing cryptocurrency from job applicants by posing as AI or blockchain companies. The report said that between December 2025 and July 2026, WaterPlum hackers infected at least 30,000 devices across 100 countries and stole funds or credentials from about 7,000 cryptocurrency wallets. The primary targets of the campaign are web designers, engineers and cryptocurrency specialists. – https://therecord.media/north-korean-hackers-infect-thousands-of-devices-waterplum-scheme

AI Helps Hackers Hijack OpenAI Staff Accounts Through a Forum – Three researchers at Hacktron just took over ChatGPT and Codex accounts belonging to OpenAI staff. The attack did not rely on phishing techniques or a leaked password. Through an image upload on OpenAI’s own help forum. The forum runs on Discourse, and until recently anyone logging in there through “Sign in with OpenAI” could, in theory, have had their ChatGPT and Codex accounts hijacked. Since staff connects all sorts of services to those accounts, GitHub, Slack, email, the actual blast radius was enormous. The whole thing, from first finding the bug to sitting inside an internal OpenAI repository, took less than 72 hours. Here’s how the attack worked. Discourse sends HEIC and HEIF image uploads to ImageMagick for processing, which uses a library called libheif to decode the files. A heap buffer overflow was found in that process. This type of memory bug can allow an attacker to read or write data outside the area the program is supposed to access. The vulnerability had actually been fixed upstream the previous year, but the fix did not have a CVE and was not backported to Debian in time. Normally, turning a bug like this into a working exploit can take weeks. This time, AI helped speed things up. The researchers first gave the task to Claude Opus 4.8, but it struggled once protections such as ASLR were enabled. After Anthropic released Opus 5, they gave it the same task in a new session. The model managed to build a working exploit against a local Mac in just three hours. – https://securityaffairs.com/199378/ai/ai-helps-hackers-hijack-openai-staff-accounts-through-a-forum.html

Brevo Supply-Chain Attack Infected Over 100,000 Websites – Brevo, formerly known as Sendinblue, is a French cloud-based marketing and customer communication platform whose clients include eBay, Louis Vuitton and Michelin. The company was first compromised on September 10, when attackers exploited a vulnerability in its SAML SSO system and gained access to 138 accounts, including one belonging to cryptocurrency hardware wallet maker Trezor. Six accounts were used to send phishing emails, while contact data was exported from 43 accounts. Brevo blocked the unauthorized access, but attackers returned four days later using a compromised, long-lived Cloudflare API key. They used the key to deploy a malicious Cloudflare Worker that injected code into Brevo’s websites and three JavaScript files embedded in customer sites. “A long-lived Cloudflare API key with full account permissions was stored in application source code and was obtained by the attacker. With it, they could create Workers, routes and DNS records on Brevo’s zones without triggering an alert.” reads a Brevo’s notice of incident. “Because the Worker rewrote responses at the edge and removed security headers such as Content-Security-Policy, our origin servers and files remained unmodified and standard integrity checks did not detect the change.” – https://securityaffairs.com/199355/hacking/brevo-supply-chain-attack-infected-over-100000-websites.html

Gyazo Data Breach Exposes 23 Million User Records – Japanese software company Helpfeel is notifying Gyazo users about a data breach that compromised 23 million user records. Attackers gained unauthorized access by exploiting a vulnerability in the service’s image upload server. “We have confirmed that approximately 23.62 million records containing data related to Gyazo users were disclosed without authorization.” notice of data breach. Gyazo is a popular cross-platform tool for capturing screenshots, GIFs and short screen recordings and instantly sharing them through links. The incident highlights the risks of vulnerabilities in systems that handle and store user-generated content. Helpfeel said this week that it recently discovered unauthorized access to its Gyazo servers. The threat actor exploited a vulnerability in the image upload server on September 11 and was able to run malicious commands. – https://securityaffairs.com/199338/data-breach/gyazo-data-breach-exposes-23-million-user-records.html

RatHat Turns Android Accessibility Into an Attack Weapon – RatHat is the new Android trojan you should know about. Zimperium researchers just published a breakdown of a strain they’ve traced to China-based operators, and what makes it different isn’t the credential theft, which is standard fare by now. It’s what it does to get there. It starts the way most mobile fraud does: a text message or a shady ad pointing to a fake app store. Once someone installs the APK, the malware doesn’t just ask for permissions and hope. “RatHat is primarily distributed through deceptive phishing sites promoted via malvertising, smishing campaigns, and third-party forums, luring victims into manually downloading malicious APKs that appear to be legitimate apps” reads the report published by Zimperium. – https://securityaffairs.com/199317/malware/rathat-turns-android-accessibility-into-an-attack-weapon.html

EU Cyber Resilience Act: the first reporting phase exposes operational and technical challenges – On 11 September 2026, a significant part of the European Union’s Cyber Resilience Act (CRA) entered its operational phase. With the entry into application of the EU ‘s CRA Article 14 on mandatory reporting obligations, manufacturers of products with digital elements made available on the EU market, including products already on the market, must now report actively exploited vulnerabilities and severe incidents affecting the security of their products. An early warning must be submitted within 24 hours of becoming aware of an actively exploited vulnerability or severe incident, followed by a more detailed notification within 72 hours. Final reporting obligations subsequently apply depending on the type of event. At the same time, ENISA launched the CRA Single Reporting Platform (SRP) as the central mechanism through which manufacturers submit notifications. The SRP is intended to improve EU-wide vulnerability intelligence and coordination. ENISA will operate the platform, while notifications are received by the designated CSIRT and shared with other relevant CSIRTs. Information from manufacturer reporting can also contribute to Europe’s broader vulnerability-management ecosystem, including the EU Known Exploited Vulnerabilities (EU KEV) Catalogue. The potential strategic benefit is therefore considerable: incidents discovered by individual manufacturers can become intelligence available to a wider European defensive community. – https://dig.watch/updates/eu-cra-reporting-phase-exposes-challenges

Israel’s AI land monitoring tools reportedly extend to West Bank – Israel has reportedly advanced land enforcement activity in the West Bank through a unit within the Israel Land Authority (ILA), which uses artificial intelligence (AI) to detect suspected land use violations. The reported use of AI in the West Bank has received criticism from Palestinian representatives and the Israeli rights group Peace Now. Peace Now noted in September 2025 that Israel’s Finance Ministry had requested 21 additional staff positions for an ILA enforcement unit operating in the West Bank. Based on the UN report published in early 2026, the unit had been established in April 2025 to identify what Israeli authorities describe as ‘illegal Palestinian construction’ in the West Bank. – https://dig.watch/updates/israel-ai-land-monitoring-in-west-bank

UN Crime Congress: Technology is transforming crime – and the response to it – Sitting in her Vienna office ahead of a pivotal UN gathering in Abu Dhabi, Monica Juma is focused on a simple reality: crime has evolved faster than many of the institutions designed to combat it. Criminal networks now move money, goods and information across borders at lightning speed, exploiting everything from new technologies to global instability. Yet the same tools that help drive illicit activity could also transform the fight against it. Technology, while creating what Ms. Juma describes as “an ecosystem of criminal profiteering with a very far-reaching footprint”, can also enable faster, smarter and more coordinated responses to threats that increasingly touch every corner of the world. Those challenges – from cyber-enabled crime to trafficking and online scams – will be at the centre of discussions in Abu Dhabi at the 15th UN Crime Congress; the world’s largest gathering focused on crime prevention and criminal justice. As the newly appointed Executive Director of the United Nations Office on Drugs and Crime (UNODC), Ms. Juma is preparing to attend the event for the first time in her current role. – https://news.un.org/en/story/2026/09/1168370

Defence, Intelligence, Warfare

US military had close call after using AI for false intelligence report, sources say – The intelligence report, circulated across the US military this spring in the midst of the war with Iran, immediately set off alarm bells: A Chinese ship in the Middle East was transporting components of a nuclear weapons program. The US military swung into action with plans to intercept the vessel, according to four sources familiar with the episode. According to two of the sources, armed members of the US military were preparing to board the ship. Military planes were in the air, one of those sources and another source familiar with the incident said. It was only just before the planned operation that officials dug deeper into the report put together by a special operations command analyst and found it had been generated with the help of artificial intelligence (AI) — and that a chatbot the analyst had used inaccurately identified the material the ship was carrying. CNN was not able to learn what the misidentified cargo was. The report, according to one of the sources, was “entirely false.” But it also “almost started a war,” the source said. Any US operation against a Chinese vessel could have risked spiraling into an armed conflict between the two nations. Across the US military and the intelligence community, officials are pushing to weave AI into nearly every facet of their work, from analyzing the huge volumes of raw intelligence the US collects and selecting targets for strikes, to more mundane applications like managing budgeting, logistics and supply chains. But the episode underscores the profound risks of using this powerful, new and relatively poorly understood technology for targeting in the middle of a war. Analysts have long feared that AI could lead to a catastrophic miscalculation if nation states are relying on poor or corrupted data — the kind of miscalculation that might lead the United States to fire on a Chinese ship based on inaccurate information. In this particular instance, the analyst queried a chatbot about some intelligence reporting on the ship’s manifest that originated with US Special Operations Command Pacific, based in Hawaii. It was not clear whether the chatbot was a commercially available one or a US government product. “The internal tools are mostly just copies of the commercial stuff wearing lipstick,” a former senior US official familiar with the AI systems used by military and intelligence analysts. – https://edition.cnn.com/2026/09/18/politics/us-military-ai-false-intelligence-china-ship

Ukraine and the Middle East teach the Indo-Pacific about the new space wars – On 14 September, US Secretary of the Air Force Troy Meink publicly announced that the US Space Force had deployed ‘space control weapons’ in response to growing threats to American and allied space systems. The announcement generated predictable criticism from Russia and China that accused Washington of accelerating an arms race in space. But that criticism misses the larger point. The competition in space did not begin with this announcement. For more than two decades, China and Russia have been developing and deploying capabilities designed to disrupt, degrade or destroy the space systems that the United States and its allies increasingly depend on. The wars in Ukraine and the Middle East show why this matters. Space is no longer simply a supporting domain. Satellite communications, navigation, imagery and other space-derived services increasingly shape what happens on the terrestrial battlefield. For Australia and its Indo-Pacific partners, the lesson should be clear: in a future regional conflict, military advantage on Earth will increasingly depend on the ability to operate – and prevail – in space. – https://www.aspistrategist.org.au/ukraine-and-the-middle-east-teach-the-indo-pacific-about-the-new-space-wars/

Frontiers

Singapore launches autonomous construction equipment pilot – Singapore’s JTC Corporation and Japan’s Kajima Corporation have launched the country’s first pilot of autonomous construction equipment, testing heavy machinery under real worksite conditions. The trial is taking place at the Bulim Autonomous Yard (BAY) in Jurong Innovation District, a JTC sandbox for validating autonomous construction equipment. Initial trials cover excavators and compactors, with the Public Sector Science & Technology Policy & Plans Office and Housing & Development Board also participating. Under the pilot, an operator assigns tasks and establishes virtual geographic boundaries, or geofences, for each machine. Sensors and AI then allow the equipment to perform assigned tasks without an operator inside the cabin or directly controlling it, while staff monitor operations remotely and can intervene when necessary. – https://dig.watch/updates/singapore-autonomous-construction-equipment-pilot

The trial is taking place at the Bulim Autonomous Yard (BAY) in Jurong Innovation District, a JTC sandbox for validating autonomous construction equipment. Initial trials cover excavators and compactors, with the Public Sector Science & Technology Policy & Plans Office and Housing & Development Board also participating. Under the pilot, an operator assigns tasks and establishes virtual geographic boundaries, or geofences, for each machine. Sensors and AI then allow the equipment to perform assigned tasks without an operator inside the cabin or directly controlling it, while staff monitor operations remotely and can intervene when necessary.

KDDI and SpaceX partner on next generation Starlink Mobile service in Japan – Japan’s KDDI has signed an agreement with SpaceX to introduce the next generation Starlink Mobile V2 service through ‘au’, its mobile network brand, with high-speed satellite data and voice calls expected to launch during 2027. The agreement was signed on 10 August and announced by KDDI on 15 September with Starlink highlighting the partnership on 17 September. The companies say the service will connect compatible smartphones directly to Starlink satellites in areas outside terrestrial mobile coverage, without requiring additional equipment. KDDI said the V2 service is intended to support higher-speed broadband communications and voice over LTE (VoLTE), expanding beyond the messaging and supported app data services currently available through au Starlink Direct. – https://dig.watch/updates/kddi-spacex-partner-on-starlink-in-japan

Japan opens call for robot software distribution platform – Japan’s New Energy and Industrial Technology Development Organization (NEDO) – a national agency – has opened a call for proposals to develop a robot software distribution platform under its programme to strengthen post-5G information and communications systems. The call opened on 17 September 2026 and closes at noon on 16 October. The project will run across fiscal years 2026 and 2027 and is open to companies, organisations, universities, and other eligible applicants in Japan. – https://dig.watch/updates/japan-robot-software-distribution-platform

Google targets data centre emissions with green steel deal – Google has signed an agreement with Swedish steel producer Stegra to support production at its planned large-scale, green hydrogen-based steel plant in Boden, Sweden, as the technology company seeks to reduce emissions associated with data centre construction. Unlike conventional steel production, which relies heavily on coal to process iron, Stegra plans to use renewable electricity to produce green hydrogen that will react with iron ore. Stegra estimates that removing coal from the process could reduce emissions by up to 95% compared with conventional blast furnace steelmaking. Google said the resulting steel is designed to meet the International Energy Agency’s definition of near-zero emissions steel. – https://dig.watch/updates/google-emissions-steel-data-centre