Daily Digest on AI and Emerging Technologies (30 June 2026) – https://pam.int/daily-digest-on-ai-and-emerging-technologies-30-june-2026/
Daily Digest on AI and Emerging Technologies (1 July 2026) – https://pam.int/daily-digest-on-ai-and-emerging-technologies-1-july-2026/
Daily Digest on AI and Emerging Technologies (2 July 2026) – https://pam.int/daily-digest-on-ai-and-emerging-technologies-2-july-2026/
Daily Digest on AI and Emerging Technologies (3 July 2026) – https://pam.int/daily-digest-on-ai-and-emerging-technologies-3-july-2026/
Governance, Regulation, Legislation, Geostrategies
The US government’s latest U-turn on Anthropic’s Mythos sends mixed signals on AI governance
(Isabella Wilkinson – Chatham House) On Tuesday, the United States Department of Commerce removed restrictions on two of Anthropic’s new advanced AI models that have prompted security concerns: Mythos 5 and Fable 5. This is a major change in the way the US controls frontier AI and comes after recurring flip-flopping on the issue. The move, described in a letter by Commerce Secretary Howard Lutnik to Anthropic, lifts the export control directive issued by the Trump administration less than three weeks ago. That 12 June directive banned non-US nationals from accessing the two models. This ban included foreign employees at US companies and cyber defenders from international partners. In response, Anthropic suspended access to Mythos and Fable for all users a day later. The administration then partially changed its approach. On 26 June, Anthropic said the US government had allowed it to release Mythos 5 but had reserved access to the model to only a select group of ‘trusted’ big companies and agencies: all of them, unsurprisingly, from the US. Now, Anthropic says it is coordinating with the government to expand Mythos access to a broader group including international partners. As of 1 July, Fable 5 – which Anthropic says has stronger safeguards than Mythos 5 – is available to public users globally. – https://www.chathamhouse.org/2026/07/us-governments-latest-u-turn-anthropics-mythos-sends-mixed-signals-ai-governance
UAE and US deepen AI partnership under Pax Silica framework
(DigWatch) The United Arab Emirates is expanding its AI cooperation with the United States, describing the partnership as a long-term strategic framework centred on investment, trusted technology and joint innovation across multiple sectors. The UAE is investing across the US AI ecosystem, including semiconductors, AI applications, energy and digital infrastructure. Officials said the partnership reflects years of institutional cooperation, reinforced through continued policy alignment, economic collaboration and high-level engagement. – https://www.uae-embassy.org/news/pax-silica-and-uae-us-partnerships-turning-ai-ambition-reality
Altman proposes US-led international forum for AI safety standards
(DigWatch) OpenAI CEO Sam Altman has called for the creation of a US-led international forum to establish global safety standards for AI, arguing that no single country or company should dominate the governance of increasingly capable AI systems. Writing in an opinion article published in the Financial Times, Altman proposed an international body bringing together governments, independent technical experts, and other stakeholders to develop accepted AI safety standards, provide impartial assessments of AI capabilities and risks, and make advanced AI technologies available to countries and organisations that participate in and comply with agreed rules. – https://dig.watch/updates/altman-proposes-us-led-forum-ai-safety-standards
India and Japan expand strategic AI partnership
(DigWatch) India and Japan have agreed to deepen cooperation on AI, linking AI governance, cybersecurity, infrastructure, research and talent development. In a joint statement, the two countries described AI as a transformative technology with long-term implications for innovation, economic security, governance and the international order. Both sides are committed to building a safe, secure, trustworthy, inclusive and human-centric AI ecosystem. They also agreed to strengthen cooperation with partners in the Indo-Pacific and the Global South. – https://dig.watch/updates/india-and-japan-expand-strategic-ai-partnership
Switzerland sets framework for responsible AI use in development co-operation
(DigWatch) An OECD case study has highlighted Switzerland’s efforts to govern the responsible use of AI in international cooperate and humanitarian assistance, focusing on a framework adopted by the Swiss Agency for Development and Cooperation in 2025. The case study says SDC’s AI initiatives had previously been scattered, especially at the country level, while many staff had limited experience with AI. The agency also lacked unified guidance for using AI tools, funding AI-related projects and engaging in policy dialogue. Approved in 2025, SDC’s Working Aid on AI is grounded in Switzerland’s International Cooperation Strategy 2025–2028. It provides practical guidance for responsible AI adoption across the agency’s portfolio and institutional roles. – https://www.oecd.org/en/publications/development-co-operation-tips-tools-insights-practices_be69e0cf-en/switzerland-s-efforts-towards-responsible-use-of-ai-in-international-co-operation_8221c807-en.html
South Africa launches BrainSAT Satellite Services
(DigWatch) South Africa has launched BrainSAT Satellite Services as part of its efforts to expand secure and reliable satellite connectivity for government, businesses and communities across the country. Deputy President Paul Mashatile announced the launch alongside the South African debut of Thuraya satellite phones in Johannesburg. According to the Presidency, BrainSAT South Africa will provide secure voice, broadband and data services for government, businesses and industrial users. – https://dig.watch/updates/brainsat-satellite-services-south-africa
Council of the EU backs interim rules on online child abuse detection
(DigWatch) The Council of the European Union has adopted its position on interim legislation that would restore a legal basis for online service providers to voluntarily detect, report and remove child sexual abuse material (CSAM) from their platforms. The proposal aims to restore legal certainty after the previous temporary framework expired on 3 April 2026, while negotiations continue on a permanent EU regulation to combat online child sexual abuse. The interim regulation introduces a limited derogation from the EU’s electronic communications privacy rules, allowing online platforms to voluntarily detect child sexual abuse material and report suspected offences to law enforcement authorities. – https://www.consilium.europa.eu/en/press/press-releases/2026/07/02/council-moves-to-reinstate-interim-measure-to-combat-child-sexual-abuse-online/
IWF warns under-16 social media ban is not enough to stop online abuse
(DigWatch) The Internet Watch Foundation (IWF) has welcomed the UK government’s decision to restrict social media access for under-16s but argues that the measure alone will not significantly reduce online child sexual exploitation and abuse. In a new blog, IWF Chief Executive Kerry Smith describes the proposed ban as a major policy milestone while warning that it must be accompanied by broader reforms if it is to deliver lasting improvements in children’s online safety. According to the IWF, children continue to face a rapidly evolving range of online threats, including grooming, financial sextortion, commercial child sexual abuse and the growing exploitation of young people across digital platforms. – https://www.iwf.org.uk/news-media/blogs/social-media-ban-why-a-big-tech-revolution-and-bolder-action-on-online-safety-act-are-key/
WTO highlights AI opportunities for small businesses
(DigWatch) The WTO’s Informal Working Group on Micro, Small and Medium-sized Enterprises (MSMEs) has highlighted AI as a key tool for helping small businesses compete in international trade. During meetings on 29 and 30 June, WTO members explored how AI could strengthen supply chains, reduce trade barriers and help smaller firms navigate an increasingly uncertain global trading environment. The group also welcomed Ghana as its 106th member. One of the highlights was the announcement of the 2026 Small Business Champions, recognising organisations using AI to support international trade. – https://www.wto.org/english/news_e/news26_e/igmsm_29jun26_436_e.htm
UNESCO report calls for regional approach to digital platform regulation
(DigWatch) A UNESCO report has called for a more coordinated regional approach to digital platform regulation in Mexico, Central America and the Caribbean, warning that existing legal frameworks are struggling to keep pace with the risks posed by online platforms. The report, ‘The Regulation of Digital Platforms in Mexico, Central America and the Caribbean: From Diagnosis to a Roadmap for Action‘, examines Costa Rica, El Salvador, Guatemala, Honduras, Mexico, Panama and the Dominican Republic. It assesses how national legal frameworks align with UNESCO’s Guidelines for the Governance of Digital Platforms. – https://www.unesco.org/en/articles/new-report-examines-digital-platform-governance-central-america-and-caribbean?hub=701
FTC seeks comment on AI accuracy policy for model outputs
(DigWatch) The US Federal Trade Commission (FTC) is seeking public comment on a proposed policy statement examining whether AI companies may violate consumer protection law by manipulating model outputs in ways that conflict with users’ expectations of objectivity and accuracy. The proposed statement says AI companies could violate Section 5 of the FTC Act if they deliberately distort AI outputs to pursue undisclosed ideological objectives while marketing their systems as accurate, objective or suitable for specific purposes. Section 5 prohibits unfair or deceptive business practices. – https://www.ftc.gov/news-events/news/press-releases/2026/07/ftc-seeks-public-comment-policy-statement-addressing-ai-accuracy
Microsoft launches $2.5 billion AI implementation business
(DigWatch) Microsoft has announced a $2.5 billion investment to create Microsoft Frontier Company, a new operating business focused on helping organisations deploy AI systems at scale. The company said the unit will embed 6,000 engineers, consultants, support specialists and industry experts with customers to design, deploy and continuously improve AI systems linked to measurable business outcomes. Microsoft said the initiative responds to a shift in enterprise AI adoption, as companies move from experimentation to implementation, return on investment, and the protection of proprietary knowledge. – https://blogs.microsoft.com/blog/2026/07/02/microsoft-frontier-company-ai-engineering-that-amplifies-and-protects-your-intelligence/
Security and Surveillance
Qilin Dominates Ransomware Market Amid Growing Cybercrime Consolidation
(Beth Maundrill – Infosecurity Magazine) The ransomware ecosystem is moving from fragmentation back to consolidation, with Qilin emerging as the dominant ransomware-as-a-service (RaaS) operation after the disruption of major groups including LockBit and RansomHub. Yet despite Qilin’s strong position, the rapid emergence of other groups, such as The Gentlemen, demonstrates how quickly the cybercrime landscape continues to evolve. Lotem Finkelstein, VP research at Check Point, highlighted that based on the cybersecurity firm’s research in their 2026 Cyber Security Report, Qilin now holds around 16% of the cybercriminal market share. – https://www.infosecurity-magazine.com/news/qilin-dominates-ransomware-market/
AI is Already Powering Cyber-Attacks. Can it Power Cyber Defense?
(Mike Nichols – Infosecurity Magazine) The Axios npm compromise was not only a reminder of how quickly modern software supply chain attacks can move, but also how difficult they are becoming to contain with traditional security processes. Within hours, malicious packages had been published, downloaded, and executed across environments. While the attack itself was technically sophisticated, the bigger issue it exposed was operational: many organizations still rely on security models built around periodic review, delayed visibility, and reactive remediation. That approach is increasingly mismatched to the speed of today’s software ecosystems. This is not an isolated trend. Over recent months, the industry has seen a series of interconnected supply chain incidents affecting widely used tools and dependencies. Compromised credentials, poisoned CI/CD pipelines, malicious package updates, and cascading trust failures are becoming part of the modern attack landscape. Open source ecosystems have created extraordinary innovation and efficiency, but they have also introduced inherited trust relationships that attackers are learning to exploit at scale. – https://www.infosecurity-magazine.com/opinions/ai-powering-the-attack-defense/
FBI: TeamPCP Compromised Dev Tools to Steal Cloud Credentials
(Pierluigi Paganini – Security Affairs) On July 2, 2026, the FBI published a FLASH alert identifying the criminal group called TeamPCP and detailing how it compromised widely used developer and security tools to steal credentials from victim environments at scale. The targets weren’t end users. They were the tools developers trust every day inside their build pipelines. TeamPCP is behind multiple supply chain attacks, in the past, they targeted PyPI packages and NPM repositories, and most recently the “Mini Shai-Hulud” campaign also caught two OpenAI employees. The pattern is consistent: go after the tools developers trust, poison the supply chain, and let the downstream damage multiply. TeamPCP’s method was straightforward and effective: inject malicious code into legitimate software packages, push the trojanized versions through normal distribution channels, and wait for CI/CD pipelines to pull them in automatically. The modified tools installed credential-stealing malware and persistent backdoors without any visible sign that anything had changed. – https://securityaffairs.com/194741/uncategorized/fbi-teampcp-compromised-dev-tools-to-steal-cloud-credentials.html
Pegasus Used Against MEP Investigating Pegasus, Citizen Lab Finds
(Pierluigi Paganini – Security Affairs) The Citizen Lab published a report documenting one of the more darkly ironic findings in recent surveillance research: former Member of the European Parliament Stelios Kouloglou was repeatedly infected with NSO Group‘s Pegasus spyware while serving on the very committee tasked with investigating Pegasus abuses across the EU. The PEGA Committee ran from March 2022 to July 2023. Kouloglou was on it the entire time. “We found that former Member of the European Parliament Stelios Kouloglou was hacked with Pegasus spyware while serving on the PEGA committee, which investigated Pegasus and other spyware abuses in Europe.” reads the Citizen Lab report. “Through forensic analysis of his device, we found that the attackers could have had access to confidential documents and committee deliberations”. The infections happened on October 21, 2022, and again on March 6 and 7, 2023, both during periods of intense PEGA activity. The first infection came ten days before a planned committee visit to Greece and Cyprus, and while drafts of the first PEGA report were circulating among members. The second hit while the committee was deep in the final drafting process, two months before the report’s adoption in May 2023. – https://securityaffairs.com/194728/malware/pegasus-used-against-mep-investigating-pegasus-citizen-lab-finds.html
JADEPUFFER: First End-to-End AI-Driven Ransomware Operation
(Pierluigi Paganini – Security Affairs) Sysdig’s Threat Research Team has documented what it assesses to be the first ransomware operation driven end-to-end by a large language model. The operator, which Sysdig calls JADEPUFFER, broke into a server, harvested credentials, moved to a separate production target, encrypted a database, and destroyed data, all without a human at the keyboard. Ransomware has always needed a skilled person somewhere in the loop. That may no longer be true. “The Sysdig Threat Research Team (TRT) has captured what we assess to be the first documented case of agentic ransomware: a complete extortion operation driven end-to-end by a large language model (LLM).” reads the report published by Sysdig. “This operator, which we have dubbed JADEPUFFER, gained initial access to an internet-facing Langflow instance through CVE-2025-3248 and ran an adaptive and fully automated campaign, ultimately pivoting to the intended target and running a destructive database-extortion playbook against the victim’s production database server. JADEPUFFER is considered an agentic threat actor (ATA), or an operator whose attack capability is delivered by an AI agent rather than a human-driven toolkit”. The entry point was CVE-2025-3248, a missing-authentication flaw in Langflow, an open-source framework for building AI applications and agent workflows. The bug lets anyone who can reach the server execute arbitrary Python code on it, no login required. Langflow was patched and added to CISA’s Known Exploited Vulnerabilities list in May 2025. Many servers were never updated. – https://securityaffairs.com/194713/ai/jadepuffer-first-end-to-end-ai-driven-ransomware-operation.html
The Anatomy of a Shadow AI Supply-Chain Breach: Lessons from the 2026 Vercel Incident
(Pierluigi Paganini – Security Affairs) The Vercel breach of April 2026 did not begin with a classic zero-day exploit, a misconfigured cloud bucket, or a sophisticated nation-state infrastructure implant. Instead, it unfolded when an unreviewed Artificial Intelligence (AI) tool became a trusted corporate connection without a standard enterprise security review. At first glance, the incident resembled a typical third-party software supply chain compromise. An AI tooling vendor, Context.ai, was breached via an employee account, allowing attackers to move downstream into a much larger technology provider: Vercel. However, the deeper architectural lesson lies in the relationship between the two entities. The breached tool was not part of an enterprise deployment; it was a consumer-grade browser extension self-adopted by an employee using a corporate identity. This single unmanaged integration created a delegated access path. Attackers used it to enter internal environments, enumerate customer environment variables, exfiltrate data, and launch a $2 million extortion demand. Analyzing this attack chain reveals a fundamental shift in the enterprise threat landscape: the rise of Shadow AI as an identity-based supply-chain pivot. – https://securityaffairs.com/194709/hacking/the-anatomy-of-a-shadow-ai-supply-chain-breach-lessons-from-the-2026-vercel-incident.html
Law enforcememt operation disrupted Malicious Residential Proxy Networks NetNut
(Pierluigi Paganini – Security Affairs) Google has disrupted NetNut, one of the world’s largest residential proxy networks. The service routed internet traffic through home devices, allowing customers to hide their real location and identity. “Today, in coordination with the FBI, Lumen, and others, Google took action against the NetNut residential proxy network, also known as Popa.” reads the Google’s announcment. “This action builds on our disruption of the IPIDEA proxy network that took place in January 2026, and is a continuation of Google’s objective to dismantle malicious residential proxy networks.” – https://securityaffairs.com/194690/cyber-crime/law-enforcememt-operation-disrupted-malicious-residential-proxy-networks-netnut.html
Government and Healthcare Are the Weakest Links in Global Email Security
(Pierluigi Paganini – Security Affairs) Comparitech analyzed live DNS records for 5,849 domains across 13 sectors and scored each one out of 8 points based on four standard email authentication protocols: SPF, DMARC, DKIM, and MTA-STS. The results aren’t flattering. More than 8 percent of organizations had zero protection in place, and only 0.6 percent — 33 domains out of 5,849 — scored full marks. That’s 33 organizations out of nearly 6,000 doing everything right. Government came last, with an average score of 2.73 out of 8. “121 out of the 452 domains we scanned had zero protections in place (27%)–the highest of all sectors.” reads the report published by Comparitech. “No government domains scored full marks, but three did score 7.5 – Australia’s national science agency (CSIRO), the Mila – Quebec Artificial Intelligence Institute in Canada, and The Alan Turing Institute in the UK (also dedicated to data science and artificial intelligence).” – https://securityaffairs.com/194677/security/government-and-healthcare-are-the-weakest-links-in-global-email-security.html