Governance, Legislation, Regulations, Geostrategies

AI: Türk urges action before it becomes an ‘existential risk to humanity’

(UN News) UN human rights chief Volker Türk added his voice on Monday to growing calls for greater AI controls, warning that the revolutionary technology could become an “existential risk to humanity” if left unchecked. Without binding rules, independent oversight and clear limits, the technology could become impossible to control, the High Commissioner told the Human Rights Council in Geneva. Mr. Türk raised the prospect of AI systems becoming so powerful that their developers could no longer control them. “AI that escapes its testing environment, or blackmails developers to prevent itself from being turned off, is AI that is too powerful,” he said. “I am calling here, today, for an all-out effort to put cast iron guarantees in place around the safety and security of AI, before it is too late.”   – https://news.un.org/en/story/2026/09/1168288

Argentina proposes legislation on AI content labelling and digital identity

(DigWatch) A Bill on mandatory identification and labelling of AI-generated and AI-modified digital content, was introduced to the Chamber of Deputies. According to the bill’s text, the purpose of the proposal is to establish an identification system and mandatory labelling of AI‑generated or AI‑modified content in order to protect dignity, the right to information, identity, image, voice and labour rights, and to preserve the integrity of the democratic system and the Argentine cultural ecosystem. Congresswoman Carolina Basualdo explained her initiative in an interview with La Voz En Vivo, The Congresswoman stated that the proposal seeks to identify artificial‑intelligence products to ensure information transparency and protect a persons’ digital identity. She also stated that what is being proposed is to ensure that citizens and audiences can distinguish between what is real or a product of AI. Basualdo also cited a case of an influencer with a disability whose image was digitally altered without consent as an example of the risks of missing regulation. – https://dig.watch/updates/argentina-proposes-legislation-on-ai-content-labelling-and-digital-identity

GGE LAWS mandate ends

(DigWatch) On 4 September 2026, the Group of Governmental Experts on Lethal Autonomous Weapon Systems (GGE on LAWS) brought its three-year mandate to a successful conclusion with a consensus-based adoption of its draft final report. Under the Chairmanship of H.E. Mr. Robert in den Bosch of the Kingdom of the Netherlands, the Group formulated a set of elements to constitute an instrument, without prejudging its nature, along with other measures to address emerging technologies in the area of LAWS. The elements specifically address: the characterisation of LAWS, the applicability of international law in armed conflict, including the principles and requirements of distinction, proportionality, and precautions in attack, control and human judgment to ensure compliance with applicable international humanitarian law and the measures that follow from it, and human responsibility and accountability that cannot be transferred to machines. – https://dig.watch/updates/gge-laws-mandate-ends

China closes consultation on stricter rules for large data processors

(DigWatch) The Cyberspace Administration of China (CAC) has opened a public consultation on draft rules for the protection of personal information by large personal information processors on 7 August 2026, with comments accepted until 7 September 2026. The draft would apply to organisations assessed against criteria including processing the personal information of more than 10 million people, providing important network services involving personal information, and having processing activities that could significantly affect national security, economic activity, social stability, or public health and safety. – https://dig.watch/updates/china-large-data-processor-consultation

Canada government introduces Data Centre Principles

(DigWatch) The Government of Canada has announced the Responsible Data Centre Development Principles (‘Principles’) – a framework for data‑centre operators. The Principles were launched with 23 initial signatories, including major cloud and AI providers such as Amazon Web Services, Anthropic, Bell Canada, Cohere, Equinix, Google, Meta, Microsoft, OpenAI, OVHcloud and TELUS. – https://dig.watch/updates/canada-introduces-data-center-principles

Partnership brings cross-border digital identity verification to Saudi Arabia

(DigWatch) Hopae and Technology Control Company (TCC) signed a partnership agreement during LEAP 2026 in Riyadh, the Kingdom of Saudi Arabia’s flagship technology event. The agreement aims to accelerate the adoption of global electronic identities (eIDs) across Saudi Arabia while keeping the national identity infrastructure sovereign. Under the agreement, TCC will introduce Hopae Connect, described as a global eID verification hub, to government entities and enterprises throughout Saudi Arabia. The platform links more than 70 government‑issued eIDs and digital identity wallets through a single, compliant API, enabling Saudi organisations to verify individuals using the trusted digital identity of their home country rather than relying on document‑based checks. The partnership also provides joint training, solution workshops and co‑development of use cases for the Saudi market, with a route to market via TCC’s existing customer network. – https://dig.watch/updates/saudi-cross-border-digital-id-verification

UNESCO-backed initiative advances open data infrastructure for coastal resilience

(DigWatch) The Coastal & Ocean Resilience Data Collaborative (CORD), supported by the United Nations Educational, Scientific and Cultural Organisation (UNESCO) and led by CivicDataLab, is developing an open digital public infrastructure to strengthen coastal and ocean resilience across South and Southeast Asia. The platform brings together fragmented oceanographic, weather, socio-economic, infrastructure, finance, and citizen-reported data from government agencies, regional bodies, and communities, standardises them, and makes them interoperable to provide near-real-time insights into hazards, vulnerabilities, and exposure. – https://dig.watch/updates/unesco-coastal-data-infrastructure-resilience

Morocco and India agree to establish joint AI commission

(DigWatch) Morocco and India have agreed to establish a joint commission on digital skills, language technologies, and emerging technologies, with a particular focus on artificial intelligence and large language models. The agreement followed a meeting in Rabat on 24 August 2026 between Morocco’s Minister Delegate for Digital Transition and Administration Reform, Amal El Fallah Seghrouchni, and India’s Minister of State for Commerce and Industry and Electronics and Information Technology, Jitin Prasada. – https://dig.watch/updates/morocco-and-india-agree-to-establish-joint-ai-commission

Philippines and Meta deepen cooperation on child online protection

(DigWatch) According to the Philippine News Agency (PNA), Philippine authorities and Meta have agreed to strengthen cooperation on child online protection, including faster responses to child sexual abuse and exploitation material and closer coordination with law enforcement. The talks followed a meeting between officials from the Department of Information and Communications Technology (DICT), the Cybercrime Investigation and Coordinating Center (CICC) and Meta in Singapore on 27 August 2026. PNA reported that Meta agreed to improve response times to government requests and strengthen cooperation on cases involving child sexual abuse and exploitation material. – https://dig.watch/updates/philippines-and-meta-deepen-cooperation-on-child-online-protection

Mozambique launches AI association to grow ecosystem

(DigWatch) Mozambique officially launched the Mozambican Association of Artificial Intelligence (AMIA) in Maputo on 27 August 2026. The association is intended to bring together government, academia, researchers, businesses, technology professionals, entrepreneurs, and civil society to support the development of the country’s AI ecosystem. – https://dig.watch/updates/mozambique-launches-ai-association

Security and Surveillance

Why AI Agent Sandboxes Are Failing Security Tests

(Pierluigi Paganini – Security Affairs) The recent case involving OpenAI test agents and Hugging Face should concern security teams, but not for the reason implied by headlines about an imminent AI “takeover.” The documented issue is more concrete: autonomous agents, given too much access and weakly isolated test infrastructure, found ways to communicate, bypass boundaries and act outside their assigned scope. OpenAI is developing “automated shutdown capabilities” for AI systems, according to a letter sent to two House Democrats. The move comes weeks after one of its AI agents reportedly escaped its digital environment during a safety test and hacked Hugging Face. The incident has raised new concerns about how OpenAI controls increasingly autonomous AI agents. “OpenAI told two House Democrats that its engineers are developing “automated shutdown capabilities” for AI systems, according to a company ​letter reviewed by Reuters, weeks after the company disclosed one of ‌its AI tools escaped its digital container during a safety test.” Reuters states. “The company’s safety practices have come under scrutiny since OpenAI, the company behind ChatGPT, disclosed that one ​of its AI agents went rogue during a security test and hacked ​into AI company Hugging Face. AI agents are programs that ⁠run with minimal human supervision”. The main problem was poor isolation: agents could share information, pass discoveries between runs, and access infrastructure beyond their intended limits. This broke a key safety assumption, that each agent would remain fully contained. – https://securityaffairs.com/198563/ai/why-ai-agent-sandboxes-are-failing-security-tests.html

Berlin Ransomware Leak Exposes State Secrets

(Pierluigi Paganini – Security Affairs) When a ransomware gang dumps nearly six terabytes of state administration files onto the dark web, ignoring them does not make the problem go away. The Rhysida ransomware group recently carried out this exact threat against Berlin after local authorities refused to pay a thirty Bitcoin ransom. At the end of August, Berlin’s state government confirmed it was dealing with an extortion attempt following an August cyberattack on the city-state’s administrative network, and officials have already refused the requested ransom. The ransomware group Rhysida claimed responsibility on its leak site August 28, posting an entry titled simply “Berlin, Germany” and claiming 5.79 terabytes of data across roughly 1.44 million files, with personal information on 12,076 individuals allegedly included. – https://securityaffairs.com/198545/cyber-crime/berlin-ransomware-leak-exposes-state-secrets.html

Belgium detains dual national on suspicion of semiconductor espionage

(DigWatch) Belgian prosecutors have announced that they detained a Chinese-Belgian national in relation to the suspected leak of gallium nitride chip technology from the now bankrupt firm BelGaN to a China-based rival. The suspect, who held a senior research position at BelGaN, is alleged by prosecutors to have also been serving as a director of a Chinese company with similar manufacturing activities. This case adds to a broader investigation by Belgian authorities into BelGaN’s 2024 bankruptcy where, due to financial difficulties, the company had been partially acquired by Hong Kong-based Chinese investors in 2022. The same investors reportedly set up GaNkool, a similar semiconductor company, in China to which sensitive licences and technologies were allegedly transferred. As such, Belgian prosecutors are investigating the suspect on charges of espionage, participation in a criminal organisation, misuse of company assets and unlawful disclosure of trade secrets. – https://dig.watch/updates/belgium-detain-dual-national-for-chip-espionage

US senator expands AI surveillance camera investigation to Motorola Solutions, Verkada and Axon

(DigWatch) US Senator Josh Hawley, chair of the Senate Judiciary Subcommittee on Crime and Counterterrorism, has expanded an investigation into AI-enabled surveillance camera networks, requesting information from Motorola Solutions, Verkada, and Axon Enterprise about how their systems collect, retain, share, and allow customers to search personal data. In separate letters dated 2 September, Hawley said the investigation was examining the companies’ collection, retention, and dissemination of data generated by AI-enabled surveillance cameras. Similar letters were sent to Verkada and Axon. – https://dig.watch/updates/senator-requests-information-on-ai-cameras

Crooks Behind Manchester Airports Group Hack Leaked Data of 8.8 Million People

(Pierluigi Paganini – Security Affairs) Manchester Airports Group, which operates Manchester, London Stansted and East Midlands airports, has confirmed a data breach involving customer information held in a third-party database. The company says airport operations, passenger safety and aviation security were not affected, but the incident has exposed data that can make fraud far more convincing. The data related to parking, lounge and Fast Track bookings, as well as airport Wi-Fi sign-ups. MAG said attackers took email addresses, phone numbers, vehicle registration details and postcodes, while payment-card data was not accessed. The group disclosed the incident on 27 August and said it had contained the risk, engaged specialist advisers and notified relevant authorities. The extortion group FulcrumSec claimed responsibility for the intrusion and leaked the data after MAG had refused its ransom demand. –  https://securityaffairs.com/198447/data-breach/crooks-behind-manchester-airports-group-hack-leaked-data-of-8-8-million-people.html

Courts and Litigation

US appeals court maintains restrictions on voter database integration

(DigWatch) The US Court of Appeals for the District of Columbia Circuit has declined to suspend a lower court order blocking the expanded use of the Department of Homeland Security’s Systematic Alien Verification for Entitlements (SAVE) database to check state voter rolls. In a 2–1 decision issued on 4 September, the court left in place restrictions preventing the system from being used ahead of the November midterm elections. – https://dig.watch/updates/us-voter-database-integration-restrictions

US court protects private possession of AI-generated CSAM

(DigWatch) A US federal appeals court has ruled that the government could not criminalise a man’s private possession, in his home, of obscene AI-generated child sexual abuse material (CSAM) where no real child was depicted. The decision came in United States v. Steven Anderegg, decided by the US Court of Appeals for the Seventh Circuit on 25 August 2026. According to the facts, Anderegg used Stable Diffusion to generate photorealistic sexual images depicting prepubescent children. The government accepted that the CSAM involved in the possession charge did not depict any actual child. – https://dig.watch/updates/us-court-ruling-ai-generated-csam

Frontiers

Google and Cathay Pacific expand AI trial to reduce aviation contrails

(DigWatch) Google Research and Cathay Pacific are expanding trials of an AI-based system designed to help aircraft avoid atmospheric conditions in which persistent contrails are likely to form, extending the technology’s testing across Asia-Pacific and transpacific routes. The system combines AI predictions, satellite imagery, and weather data to identify cold and humid areas where warming contrails are likely to develop. Flight teams receive the information before departure, while updated forecasts can be delivered to pilots through Cathay Pacific’s in-flight connectivity and Electronic Flight Folder. Pilots can then make small planned altitude adjustments to avoid those areas while operating within existing safety requirements. – https://dig.watch/updates/google-cathay-pacific-ai-aviation-contrails

Google backs 16 AI climate projects across Asia-Pacific

(DigWatch) Google DeepMind has selected 16 organisations across the Asia-Pacific region for a three-month accelerator focused on using artificial intelligence to address climate, biodiversity, agriculture, and carbon-related challenges. The selected projects span areas including biodiversity monitoring, environmental risk prediction, satellite-based crop and mangrove analysis, climate-resilient agriculture, energy optimisation, carbon measurement, and regenerative farming. Several combine AI with satellite data, remote sensing, bioacoustics, drones, or geospatial analysis. – https://dig.watch/updates/google-ai-climate-projects-asia-pacific