Governance, Regulation, Legislation, Geostrategies

 

How governments approached AI and digital issues at UNGA81 – AI and digital technologies emerged across discussions of international security, development, sovereignty, inequality and human rights during the UN General Assembly’s 81st session, with governments outlining different priorities and approaches to governing technologies that are advancing faster than existing institutions and rules. DigWatch’s full report on digital and AI issues at the UNGA81 General Debate brings together statements tracked throughout the session, moving beyond the daily coverage to map the major themes and the positions expressed by individual countries. AI governance featured prominently, but the discussion extended well beyond regulation. Speakers addressed who controls the technologies and resources underpinning AI, whether developing countries will participate in building AI or remain consumers of systems developed elsewhere, how human control can be preserved as AI capabilities grow, and what role the UN should play in international governance. Technology also featured in discussions of international peace and security. Speakers raised issues ranging from cyberattacks, disinformation and threats to critical infrastructure to autonomous weapons and the military use of AI. Some described technological competition itself as an arena of geopolitical rivalry, alongside trade wars and hybrid attacks. At the same time, proposals were put forward for the use of AI in diplomacy, mediation and conflict prevention. One proposal called for an international code, developed under UN auspices, for the peaceful use of AI in international relations and diplomacy. Other speakers called for greater regulation of AI in warfare and international rules governing autonomous weapons. Autonomous weapons prompted particularly strong calls for human control and accountability. Speakers argued that lethal force and military uses of AI, including in nuclear command and control, must remain under human judgement, accountability and international law. Calls included a legally binding instrument containing prohibitions and restrictions on autonomous weapons. – https://dig.watch/updates/governments-approach-ai-digital-issues-unga81

 

Google invests $1.2 million in water conservation near Chilean data centre – Google has announced a $1.2 million contribution to improve water conservation in Chile’s Maipo River Basin, where its Quilicura data centre operates. The funding will support Aguas Andinas in upgrading a 2.5-kilometre section of the Unidos de Buin Canal Association’s main irrigation canal, which supplies water to local agriculture. The project in Chile will replace the canal’s existing earthen surface with concrete lining to reduce water lost through seepage. The improvements are also intended to limit sedimentation and erosion, improving water quality. Google estimates that the upgraded infrastructure could save up to 7.4 million cubic metres of water annually over its lifespan. – https://dig.watch/updates/google-water-conservation-chilean-data-centre

 

China plans big data centre for national water network – China has outlined plans to modernise its national water network during 2026–2030, including new digital infrastructure to improve flood forecasting, water allocation, and environmental protection. Minister of Water Resources Li Guoying announced the plans on 30 September, with the government targeting water network coverage of 82% of the country’s territory by 2030. The government plans to establish a national water network dispatch centre, a big data centre, and basin-level sub-centres. These facilities are intended to support comprehensive monitoring, intelligent analysis, forecasting, and decision-making across water management systems. – https://dig.watch/updates/china-data-centre-national-water-network

 

UK restores Union Learning Fund to support digital and AI skills – The UK government has announced £15 million a year in additional funding to restore the Union Learning Fund in England, supporting workers in developing digital skills and adapting to changes in the workplace, including those brought by AI.   The fund, which previously generated around 180,000 learning opportunities annually, will work through trade unions and employers to provide training ranging from essential English, maths, and digital skills to skills for growing industries such as advanced manufacturing. Employees will not need to be union members to access the training. – https://dig.watch/updates/uk-union-learning-fund-support-digital-and-ai

 

Security and Surveillance

 

UN and Viet Nam spark the UNODC Regional Counter-Cybercrime Hub for Asia and the Pacific, to open in 2027 – The United Nations and Viet Nam have advanced plans for a UNODC Regional Counter-Cybercrime Hub for Asia and the Pacific, to be located in Ha Noi, following a Letter of Intent signed between Viet Nam’s Ministry of Public Security, General Luong Tam Quang, and the United Nations Office on Drugs and Crime (UNODC) Executive Director Monica Juma, on the margins of the UN General Assembly High-Level Week in New York. “The Letter of Intent gives practical shape to what Viet Nam began when it hosted the signing of the UN Convention against Cybercrime in 2025. Viet Nam is showing a strong commitment to fighting cybercrime, and work like this will help make the region, and the world, a safer place. The UNODC is proud to support Member States to eradicate all types of cybercrime.”, said Dr. Joshua James, Regional Counter-Cybercrime Coordinator, UNODC Regional Office for Southeast Asia and the Pacific. The agreement, exchanged in New York on 22 September 2026, provides a further institutional step towards establishing the Hub as a regional platform for cooperation against cybercrime. The initiative builds on the groundbreaking of its headquarters at the Hoa Lac Hi-Tech Park in Ha Noi in July and forms part of Viet Nam’s broader engagement with the UN on digital security and international criminal justice. – https://dig.watch/updates/un-vietnam-cybercrime-asia-pacific-hub-2027

UK accuses China of using academics to spy on AI and other tech research – Britain’s MI5 spy agency issued an unprecedented alert on ​Wednesday to warn that a Chinese organization was involved in funding academic research into topics such as AI on behalf ‌of Beijing’s state security service. In its “Espionage Alert”, MI5 said more than 100 UK-based academics had contributed to projects funded by the China General Technology Research Institute (CGTRI), which had “very strong ties” to the Chinese Ministry of State Security (MSS), Beijing’s civilian intelligence service. “MI5 has identified that the primary purpose of CGTRI is to fund academic research that directly ​improves MSS technical capability for espionage,” the alert said. “UK academics have contributed to CGTRI-funded projects on topics including Artificial Intelligence (AI), cybersecurity, covert ​communications systems and steganography. This activity supports MSS espionage, which poses a threat to UK national security.” – https://edition.cnn.com/2026/09/30/china/uk-mi5-china-spy-warning-intl-hnk

China-Linked Hackers Impersonate AI Experts to Target US Policy Insiders – A China-aligned group has posed as well-known AI policy figures and economists to steal login credentials from AI policy specialists at US think tanks, universities and law firms. Proofpoint, which tracks the group as TA419, said in research published October 1 that it has been running credential phishing against staff at think tanks, defense contractors, universities and law firms in the US and Japan since at least April 2025. It is the first public report on TA419. From July 8, the attackers sent emails in the name of Lynne Parker, who served as principal deputy director of the White House Office of Science and Technology Policy, and later economist and foreign policy expert Heidi Crebo-Rediker. In February, TA419 had used the identity of a senior Anthropic employee to contact a think-tank analyst working on AI policy. The opening messages were harmless invitations to sit on a made-up “AI Policy Advisory Committee” or help with a Senate Committee on Foreign Relations report on AI export controls. Anyone who responded got a shortened link that bounced through several redirects to a spoofed OneDrive login page. – https://www.infosecurity-magazine.com/news/ta419-impersonates-ai-experts-us/

CloudSyncD MacOS Backdoor Hides Behind Fake Zoom Installer – A new macOS backdoor has been hidden inside a fake Zoom installer that prompts users for their login password before launching an embedded second stage. The malware, dubbed CloudSyncD, has moved from development testing to samples configured against live command-and-control (C2) infrastructure. Jamf Threat Labs first encountered CloudSyncD on September 15 in a build that was still under development, according to research published September 30. Two days later, it found samples configured against live C2 infrastructure across more than one domain, indicating the operation had progressed toward deployment. – https://www.infosecurity-magazine.com/news/cloudsyncd-macos-backdoor-fake/

AI Threats Top Cybersecurity Preparedness Gap, PwC Finds – Threats targeting AI systems is the area that cybersecurity leaders currently feel last able to address, with skills, accountability and data protection gaps also looming large, according to PwC. The consulting giant polled 3934 business and tech leaders across 71 countries for its 2027 Global Digital Trust Insights report, published on October 1. Over half (52%) said adversarial AI attacks represent the biggest cyber preparedness gap. The challenge of responding to these risks is compounded by governance issues. – https://www.infosecurity-magazine.com/news/mitigating-adversarial-ai-top/

From Shadow AI to Accountable Agents: Why Agent Governance Needs Enforcement – Identity governance was already becoming more complex as human and machine identities proliferated across enterprise environments. The rise of agentic AI has raised the stakes further, introducing autonomous agents that can access data, interact with systems and take action independently. For organizations in sectors such as government, financial services and critical infrastructure, identity is increasingly becoming the control plane for AI, making effective governance essential to secure adoption. Current approaches to identity management are designed around human accounts and are not equipped to deal with fast-moving and highly adaptable AI agents. The nature of AI agents requires a fresh approach to identity governance. – https://www.infosecurity-magazine.com/blogs/why-agent-governance-needs/

AI Agent Chains Zammad Zero-Days To Take Over DIVD Systems in Seconds – The Dutch Institute for Vulnerability Disclosure, a nonprofit organization of volunteer security researchers whose whole job is finding and responsibly disclosing vulnerabilities in other people’s software, just disclosed that it got breached through two zero-days in its own ticketing system. The attackers got in through Zammad, an open-source helpdesk platform that DIVD used internally. Working with Merlon Security, DIVD identified two previously unknown vulnerabilities, tracked as CVE-2026-102489 and CVE-2026-102490. Each vulnerability was serious on its own, but chaining them made the attack much more dangerous. The attackers could hijack sessions, run code remotely and move from a regular Zammad account to root access within seconds. DIVD said the speed was linked to the use of an AI agent during the attack. “When hackers get hacked, we deal with it in hacker style. While trying to figure out how the attackers got into our own systems, we found two zero.day vulnerabilities in Zammad”, the organization wrote on LinkedIn. “Used together, they allowed the attackers to hijack sessions, run code remotely and escalate privileges from the Zammad user to root, in seconds, due to the agentic part of this hack. From there they were able to access other services and read and exfiltrate data. We urge everyone using any version of Zammad to update to version 7 or take it offline as soon as possible.” – https://securityaffairs.com/200126/hacking/ai-agent-chains-zammad-zero-days-to-take-over-divd-systems-in-seconds.html

WatchGuard fixes critical Fireware OS flaw allowing remote code execution – WatchGuard has released security updates for Fireware OS that address 15 vulnerabilities, including a critical code injection flaw, tracked as CVE-2026-86131 (CVSS score of 9.2), that could allow an attacker to execute commands with root privileges on a vulnerable Firebox. “A code injection vulnerability in WatchGuard Fireware OS’s BOVPN Over TLS client configuration handling allows an attacker who controls the remote VPN server to execute arbitrary commands as root on the connecting Firebox.” reads the advisory. The flaw affects the way Fireware OS handles configurations for BOVPN over TLS clients, a feature used to create VPN tunnels between WatchGuard Firebox appliances. The attack requires the threat actor to control the remote VPN server to which the vulnerable Firebox connects. If successfully exploited, the attacker can execute arbitrary commands as root on the connecting appliance. The vulnerability does not require user interaction or prior privileges. – https://securityaffairs.com/200108/security/watchguard-fixes-critical-fireware-os-flaw-allowing-remote-code-execution.html

 

Defense, Military, Warfare

Hegseth announces new autonomous weapons command while repeating attacks on ‘woke’ culture during speech to troops – Secretary of Defense Pete Hegseth announced a new Pentagon autonomous warfare command in a speech on Wednesday to troops at the military base in Quantico, Virginia, a sequel to the speech he delivered last year to senior officers that focused on fitness and culture. Hegseth said the new Autonomous Warfare Command, which will focus on AI robotics and autonomous weapons, will be led by a four-star general, a sign that it will have a key role in future military operations. “The autonomous and robotic revolution has enormous implications for equipment, doctrine, organization, and the assumptions that have governed warfare for decades,” Hegseth said. “Lasting change must be anchored in military command structures”. The Pentagon chief cited lessons learned from the war in Ukraine as motivation for the new command and said that a complementary effort known as Project Agincourt will work to clear the way for the new command and overhaul how the military acquires, tests, and fields drones and other similar weapons. A memo released by the Pentagon said that the new command aims to go online in October 2027. The speech also included Hegseth repeating a number of his past criticisms of military culture and saying that he had ordered further cuts to the ranks of senior officers to continue to remove “woke” policy. He doubled the percentage target for cutting general and flag officer positions in the ranks to 20%, up from a 10% reduction of generals and admirals he had directed in a May 2025 memo. Hegseth has fired more than two dozen senior officers across the military — many with little to no explanation, to the dismay of lawmakers on both sides of the aisle. The military services have also reduced the number of available four-star jobs, forcing some senior officers to leave the military. Frustration over the firings was in part responsible for the resignation of former Army Secretary Dan Driscoll. CNN previously reported that Driscoll “brought up concerns with the administration surrounding Army transformation and readiness and Hegseth’s blocking those efforts specifically by firing the generals who were responsible,” in his resignation. Some of the speech’s attendees included Tesla and SpaceX CEO Elon Musk, former Speaker of the House Newt Gingrich, and Palmer Luckey, founder and head of defense tech company Anduril. Hegseth announced that the three will chair a new future of warfare study group known as Project Meridian. – https://edition.cnn.com/2026/09/30/politics/hegseth-speaks-to-military

Frontiers and Markets

Google launches Gemini 4 Argon with restricted access for trusted cyber defenders – Google has announced Gemini 4 Argon, a new frontier AI model designed for complex, long-horizon tasks across software engineering, enterprise knowledge work and cybersecurity defence. Unlike conventional model launches, Google is initially restricting access to a selected group of trusted cyber defenders through its Fairwind Programme, reflecting the growing security implications of increasingly capable AI systems. Google says the phased approach is intended to allow early users to test the model, identify vulnerabilities and improve safeguards before broader availability. Argon’s cybersecurity capabilities are central to the launch. Google says the model has been trained to autonomously identify, validate and patch software vulnerabilities, positioning it as a tool for defensive cybersecurity teams rather than initially as a general-purpose system available to all users. For trusted defenders and Google’s own internal teams, Argon is being made available without cyber-specific guardrails so that its full defensive capabilities can be tested in controlled environments. This creates a deliberate asymmetry: actors considered capable of managing the risks receive access to capabilities that may not yet be appropriate for unrestricted public deployment. – https://dig.watch/updates/google-gemini-4-argon-restricted-access