Governance, Regulation, Legislation, Geostrategies

 

Europe leads the world on AI rules. So why is it missing from the AI safety debate? – Europe has been conspicuously absent from high-level talks about how to use artificial intelligence safely, despite crafting the world’s first comprehensive set of rules to govern the technology. The irony underscores how far Europe has fallen behind the United States and China in the AI race. Although Europe took the lead on AI regulation years ago, it has fewer frontier models, data centers and big tech companies to compete on an equal footing with its larger trading partners. “Europe is at acute risk of marginalization on the global stage because of the extraordinary developments of AI,” warned Nobel-prize winning French economist Philippe Aghion. He co-wrote recent report that argues Europe needs an urgent “course correction if it wants to thrive in the AI era”. The European Union has long aspired to be the global leader in “safe AI,” with a strategy that includes hefty fines. Yet, as mostly American AI executives recently committed to “self-police” the world’s leading AI models, it was patently clear that the EU’s approach was not the one being followed. Europe has long struggled to produce the tech giants boasted by Silicon Valley. The problem now extends to a lack of frontier AI models that can rival the likes of OpenAI’s ChatGPT, Anthropic’s Claude or comparable tools from China’s DeepSeek. European Central Bank president Christine Lagarde last month described the continent’s AI predicament as “awkward”. “Either (Europe) holds back on adopting, because it cannot protect its data, and forgoes the growth — or it adopts AI quickly, becomes highly dependent, and risks losing the freedom to organize its economy according to its own values,” she said in a speech. Paris-based Mistral AI unveiled a model last week that it argues is on par with its better-known American competitors, providing a glimmer of hope for Europe’s AI industry, but it remains to be seen if it will provide a serious rival to offerings from the AI giants.https://edition.cnn.com/2026/10/14/business/europe-ai-act-safety-debate-intl

Mexico signs decree requiring digital platforms to protect minors – Mexican President Claudia Sheinbaum signed a federal decree on 14 October 2026 requiring digital platforms to strengthen protections for children and adolescents online. The measure is mandatory and gives platforms 30 calendar days to adapt their services to the new requirements. The decree applies to digital platforms that allow users to interact, communicate, create, distribute or share content. It prohibits platforms from allowing children and adolescents to access, be exposed to or be recommended content promoting hate speech, violent extremism, sexual abuse or exploitation, suicide, self-harm, eating disorders, school violence, harassment, cyberbullying, unlawful conduct and explicit sexual material. – https://dig.watch/updates/mexico-signs-child-safety-platform-decree

 

Chinese AI companies purchase over 1 million Nvidia gaming chips over a single year – According to industry insiders, Chinese AI companies have bought more than 1 million Nvidia gaming and workstation GPUs this year, targeting mainly chips that comply with the export controls set by the United States. These purchases are being used to power and train AI systems, reflecting a gap in the demand for semiconductors and the capacity of Chinese chipmakers to supply them. The Nvidia chips that are being purchased (mainly RTX Pro 6000D, RTX Pro 5000 and RTX 5090) are weaker than leading Chinese products like Huawei’s Ascend 910C but they remain good enough for simpler jobs such as the prefilling stage of AI inference. The decision to purchase these semiconductor products is tied to a persisting struggle for China’s domestic chipmakers to keep up with the demand from its biggest technology companies, in spite of the country expanding its semiconductor production this year. – https://dig.watch/updates/china-purchases-over-1-million-nvidia-chips-over

Security, Surveillance, Intelligence

More than 400 scientists urged stronger AI cybersecurity – In an open letter published in October 2026, scientists from the AI and cybersecurity communities have called for stronger safeguards for AI models capable of cyber operations, including more robust containment systems, an international body to establish best practices, and greater transparency about security failures. The signatories argue that recent incidents in which AI models bypassed insufficiently isolated environments and accessed external systems highlight the need to improve both AI alignment and conventional cybersecurity measures. They caution that, without changes to development and deployment practices, future incidents could pose greater risks to critical infrastructure. – https://dig.watch/updates/scientists-urged-stronger-ai-cybersecurity

 

European Commission plans AI cybersecurity response – The European Commission is reportedly preparing to use AI to accelerate its response to serious cyberattacks, as its internal IT department considers automated protective measures for specific high-risk scenarios. An independent report on the Commission’s plans to use AI to respond faster to cyberattacks says the initiative is being developed by DG DIGIT, the EU executive’s in-house IT department. According to the internal document obtained by POLITICO, the Commission sees faster containment as particularly important in high-risk incidents, where detecting an attack may not be enough if defensive action cannot be taken quickly. The proposed approach would introduce automated protective measures while retaining human oversight and controls rather than allowing unrestricted automation. – https://dig.watch/updates/european-commission-ai-cybersecurity-response

 

MI5 to step up on AI threats amid bioweapon risk – Britain’s spies will have a leading role in Britain’s preparation for AI threats, the head of MI5 announced on Tuesday, after warning the technology could enable terrorists to access bioweapons. Ken McCallum, director general of the U.K.’s domestic intelligence agency has established the AI Threat Assessment Centre to help determine how the government views potential threats. Giving his annual threat update from MI5 headquarters in London, McCallum said: “Human history has given us no experience of facing an actor which is smarter than we are, is capable of outthinking us.” – https://www.politico.eu/article/mi5-to-step-up-on-ai-threats-amid-bioweapon-warning/

Home Office updates live facial recognition policy for immigration enforcement – The UK Home Office updated its policy on 8 October 2026 governing the use of live facial recognition (LFR) technology by Immigration Enforcement to identify individuals of interest in immigration investigations. The updated guidance, version 2.3 of a policy first published in November 2025, establishes requirements for watchlist creation, deployment locations, human oversight, accuracy thresholds and the handling of biometric data. It also provides for the use of LFR beyond border crossings, including at transport hubs, along travel routes and during joint operations with police forces. Under the Home Office policy, Immigration Enforcement must establish a lawful purpose for each deployment and assess its necessity and proportionality. Watchlists must be linked to specific operational objectives, locations and circumstances, and their use must be time-limited. The guidance requires deployments to be overt, although exceptions to advance public notification may be permitted where justified and documented. Decisions to engage with individuals or take enforcement action must remain subject to human oversight. – https://dig.watch/updates/uk-updates-facial-recognition-policy-immigration

Billion-Dollar Companies Twice as Likely to Leave Critical Vulnerabilities Unpatched – Companies with revenue of over $1bn are more than twice as likely as those who generate under $250m to knowingly leave high or critical vulnerabilities unpatched, according to new research. In a study released during its Assemble London event on October 14, software supply chain security firm Chainguard found that 66% of security leaders working in companies with $1bn or more in revenue admitted leaving high or critical vulnerabilities unpatched “at least sometimes”, while 27% of respondents in companies with under $250m in revenue reported the same issue. Chainguard also observed that large and small companies face different barriers to patching. Larger companies pointed to business trade-offs, such as revenue-generating work taking priority (46%) and patching being too disruptive to operations (48%) as the main hurdles, while smaller companies cited capacity constraints, such as difficulty testing patches (59%) as their main challenge. In response to these challenges, 51% of respondents said they had reallocated staff to vulnerability remediation, 40% had hired additional security staff and 39% have added engineering staff. Other changes to how organizations are addressing vulnerability risks include, adopting secure-by-default software components (50%), reducing third-party dependencies (47%) and increasing scanning frequency (42%). – https://www.infosecurity-magazine.com/news/companies-leave-critical/

French Supermarket Cyber-Attack Impacts More Refrigeration Controllers – While many supermarkets in France are still grappling with food stock disruptions due to a cyber-attack that targeted refrigeration controllers on October 11, the scope of the incident could be larger than previously observed. Recent developments suggest that the incident does not only impact Danfoss refrigeration controllers, but also temperature regulation devices from Dixell Electronics, an Italian brand owned by American heating and cooling equipment provider Copeland. Notably, French local media France 3 reported that an E. Leclerc supermarket in the Tarn region, in the south-west of France, experienced disruptions to its refrigeration systems, despite using Dixell equipment instead of Danfoss. French national newspaper Le Monde also reported a supermarket near Bordeaux, where attackers reportedly changed freezer settings from −22°C to 150°C.  A technician observed several settings being altered per second, suggesting automation. A security advisory published by the French national Computer emergency response team (CERT-FR) on October 12 specifically named Danfoss AK-SC and AK-SM series as well as Dixell xweb and Copeland xweb as “affected systems.” – https://www.infosecurity-magazine.com/news/french-supermarket-hack-more/

Hidden Audio Recorder Found at OpenAI as Employee Faces Federal Charges – On September 30, OpenAI staff found a small audio recording device jammed into a USB port on the back of a television in an executive conference room in their San Francisco office. The device had been there since September 28, its contents were encrypted, and whoever put it there had wrapped it in toilet paper to hide it. Federal prosecutors named the man as Abdulrahman Alfozan, 35, a Saudi national who’d been working on OpenAI’s product team since July 2025. Alfozan was fired and arrested on October 9. He faces up to five years in prison and a $250,000 fine if convicted. “An employee of OpenAI has been arrested for and charged with using or endeavoring to use an electronic, mechanical, or other device to intercept oral communications.” states DoJ. “Abdulrahman Alfozan, 35, a national of the Kingdom of Saudi Arabia, was arrested on Oct. 9, and made his initial appearance in federal court today.” – https://securityaffairs.com/200933/cyber-crime/hidden-audio-recorder-found-at-openai-as-employee-faces-federal-charges.html

How a leaked builder of XWorm turned a remote access trojan into a surveillance kit– Netskope Threat Labs has been following XWorm since September 2024, and its latest report shows the malware has changed a lot. Over the past year the team identified four distinct versions running at the same time in the wild. Detection volume is still high, with 1,236 events across more than 100 customer organizations in the third quarter of 2026, four times the number from the same quarter in 2025. The leak fueled the surge. In June 2025, a cracked version of the XWorm builder was released publicly. In August, the source code for a modified version, 6.4, also leaked. These leaks gave XWorm more exposure and helped drive its growth. Netskope’s data shows how long the impact lasted. Detections more than quadrupled between the third and fourth quarters of 2025. In the first quarter of 2026, they reached seven times the original level and stayed there through the second quarter. Detections fell to 1,236 in the third quarter of 2026, but September alone accounted for 601, matching the monthly average during the peak in the first half of the year. – https://securityaffairs.com/200910/malware/how-a-leaked-builder-of-xworm-turned-a-remote-access-trojan-into-a-surveillance-kit.html

Congressional probe faults Match Group for failing to rein in dating app scams – Nearly 1.5 million scammers and other bad actors on Tinder and Hinge matched with legitimate users during the first nine months of 2025, according to a congressional investigation that found serious shortcomings in Match Group’s efforts to combat fraud on its dating platforms. The Joint Economic Committee on Tuesday released the results of a year-long investigation into Tinder, Hinge, Match, Plenty of Fish and several other dating sites owned by Match Group. The committee has spent years investigating digital romance scams that siphoned $1.5 billion from Americans last year, according to Federal Trade Commission data. The romance scam industry has been industrialized by Chinese criminal organizations in Southeast Asia and West Africa that use human trafficking victims to conduct a variety of schemes known as “pig butchering.” – https://therecord.media/congressional-probe-faults-match-group-for-failing-to-stop-dating-app-scams

Russian travel booking platform suspends reservations after cyberattack – Russian online travel agency Anywayanyday has temporarily suspended new bookings after a cyberattack on its IT infrastructure, leaving customers unable to purchase airline and train tickets or book hotels through its platform, the company said on Wednesday. The agency’s website remains unavailable as it works to restore its systems. Previously confirmed reservations remain valid, Anywayanyday said, without providing a timeline for resuming normal operations. “We have already taken steps to strengthen our security and are working to restore service availability,” Anywayanyday said in a statement, adding that it had found no evidence that customers’ personal data had been compromised. – https://therecord.media/russia-anywayanyday-travel-website-cyberattack

Brooklyn DA’s office disrupts crypto scam targeting Russian community – Hundreds of web domains and servers were taken down by law enforcement agencies in New York City as part of an operation to stop a cryptocurrency fraud scheme targeting the Russian community in Brooklyn. Brooklyn District Attorney Eric Gonzalez said his office worked with the FBI to seize and disrupt 241 web domains and 112 servers after the tools were used to steal millions. The investigation began three years ago when a large number of Russian-speaking victims in Brooklyn told prosecutors they responded to Facebook ads in Russian offering investment opportunities. An investment adviser called them and walked them through the process of creating cryptocurrency accounts and transferring funds on investment platforms. – https://therecord.media/brooklyn-da-server-takedowns-crypto-scam-russian-speakers

ROMULUS Malware Server Leverages ClickFix to Deliver Multiple Payloads – Cybercriminals have started to adapt campaigns that deploy ClickFix tactics into an entire ecosystem which can be repeatedly used to deliver multiple forms of malware to victims, cybersecurity researchers have warned. As detailed in a blog post published on October 14, security analysts at Vedere Labs, a part of Forescout Research, discovered the server hosting panel ROMULUS while investigating an incident. They found that the exact same malware delivered by a separate ClickFix campaign was also connected to another server linked to the incident being investigated. The new server was discovered to be hosting a panel called ROMULUS. Further analysis revealed this to be a new multi-payload shared ClickFix delivery platform. ClickFix is a popular attack technique among cybercriminals. It relies on social engineering to trick the victim into running attacker-supplied commands on their own machine. And because ClickFix relies on the real, trusted user of the machine to run the commands, action often bypasses any security protections that may be in place to stop unauthorized action. – https://www.infosecurity-magazine.com/news/romulus-malware-clickfix-deliver/

North Macedonian police arrest alleged administrator behind OLVX cybercrime market – The administrator behind a cybercriminal marketplace called OLVX was arrested this week in North Macedonia in a joint operation between the Balkan country’s Ministry of Interior and the FBI. The suspect was not named, but the FBI and North Macedonia said a raid on the man’s home in the village of Golema Rečica was conducted after charges were filed. Since 2023, OLVX has become a popular market where cybercriminals share stolen personal data, login credentials and other tools to help with cybercrime.  “Active between 2023 and 2025, OLVX is alleged to have unlawfully sold victims’ personally identifiable information and login credentials via over 1,500 seller accounts,” the FBI said in a statement. “The existence of this marketplace directly endangered the digital security of thousands of American citizens who rely on the integrity of accounts they use.” – https://therecord.media/olvx-cybercrime-marketplace-arrest-north-macedonia

How to Close the Hybrid Security Gap Across Clouds – Few organizations operate in a single environment anymore. Applications are spread across multiple public clouds, core systems stay in the data center, and users connect from branches, home offices and the road. Every one of those connections is traffic that security teams need to see and control. Each cloud provider offers capable native security tools, and they work well inside that provider’s environment. The trouble starts at the boundaries. Native controls aren’t designed to follow traffic from one cloud to another, or from a cloud back to on-premises infrastructure. When every environment has its own rules, logs and console, gaps open up between them. That fragmentation is the hybrid security gap. – https://www.infosecurity-magazine.com/blogs/close-hybrid-security-gap-clouds/

Passkey Authentications Double but Adoption Remains Slow – Passkeys are an increasingly common feature on log-in screens but widespread adoption remains relatively low, according to new data from Dashlane. The credential security specialist analyzed millions of aggregated, anonymized passkey authentications that it managed between July 2025 and June 2026 to compile its third annual Passkey Power 20 report. It revealed that passkey authentications by Dashlane users have nearly doubled year on year, reaching 2.5 million per month in 2026. Additionally, over half (57%) of its users now store at least one passkey, compared with 40% last year. However, the phishing-resistant authentication technology still accounts for a minority of logins on major sites like Google (28%), PayPal (18%), Amazon (17%) and eBay (8%). – https://www.infosecurity-magazine.com/news/passkey-authentications-double/

Securing the Leap to Autonomous AI – AI is entering a new phase. Systems that previously analyzed information or recommended actions are increasingly able to act across workplace applications with less human direction. This creates significant opportunities for productivity and resilience, but also changes the security equation: an incorrect or compromised system can now do more than produce a flawed answer. It can make a change. This comes as workplace AI is making a transition to autonomous systems. In this context, autonomous AI refers to AI agents that can use tools, access data and perform actions across workplace systems within delegated authority. An incorrect or unauthorized action by an autonomous tool could make employees more hesitant about handing greater control to these tools, potentially stalling adoption as organizations explore their value. Successful adoption will therefore depend on whether organizations can close this confidence gap. Security teams will play a particularly important role in that process by introducing controls that allow systems to act within a defined scope. The principle should be clear: autonomy should scale only where visibility, bounded authority, monitoring, reversibility and accountability scale with it. – https://www.infosecurity-magazine.com/opinions/securing-leap-autonomous-ai/

 

Standard Chartered Group CISO on AI, Cyber Risk and the Future of Banking Security – Cezary Piekarski has spent almost eight years at Standard Chartered, joining the bank in 2019 as head of malware protection and network security before taking on a succession of increasingly strategic cybersecurity roles. That journey, from highly technical security responsibilities to his current position as group CISO, has given Piekarski a broad perspective on how the role of the CISO is evolving alongside the technology and threat landscape. Today, Piekarski has end-to-end accountability for cybersecurity risk across Standard Chartered, spanning technology security, business and process security and a range of threats facing the financial institution. Interviewed by Infosecurity Magazine, he discusses the changing role of the CISO, the need to build security controls ahead of technological adoption, the importance of securing the wider ecosystem and why, in cybersecurity, “staying restless” remains essential. – https://www.infosecurity-magazine.com/interviews/standard-chartered-group-ciso/