Cybersecurity: Israel as the shield of democracies (keio university, 27 October 2026). Last year, ransomware forced Asahi to suspend domestic orders and shipments. An attack on digital systems disrupted the physical economy. Recent Japanese reports of AI-enabled attacks underline how quickly this threat is evolving. Cybersecurity is now part of the defence of freedom. Let me make three arguments. First, democracies are prime targets of cyber aggression. Microsoft's 2026 report places Israel second and Japan fourth in its sample of observed nation-state cyber activity. It recorded 210 events against Israel and 172 against Japan. Peer-reviewed research covering 2008 to 2016 recorded roughly five times as many weekly denial-of-service attacks against domestic hosts in democracies as in autocracies. The European Central Bank finds that a small cluster of countries, often authoritarian, accounts for most recorded state-sponsored attacks. These attacks threaten essential services and sovereign decisions. Japan and Israel share a strategic challenge. Second, democratic solidarity must produce practical protection. A compromised supplier can expose partners abroad. Warnings must reach the next potential victim in time. Japan's move towards active cyber defence makes cooperation especially timely. Navon need trusted information sharing, trained people and joint exercises that test how hospitals and factories can recover. Such cooperation already delivers results. In 2022, America's Treasury reported that cooperation with Israel was providing real-time threat information to help protect America's financial sector. Third, Israel is an indispensable partner in this common defence. Close cooperation among academia, government, industry and defence turns operational experience into innovation. This March, Google completed its acquisition of Israeli-founded Wiz, following a thirty-two-billion-dollar agreement. Google reports that Wiz serves half the Fortune 100. Takeda is also a customer. According to YL Ventures, Israeli cybersecurity companies raised 5.1 billion dollars across 136 funding rounds last year. Wiz researchers also helped identify a critical vulnerability in NVIDIA's AI infrastructure software, which NVIDIA fixed. Japan's industrial expertise and Israel's cybersecurity capabilities are complementary. Partnerships are already taking shape. Yokogawa works with Claroty on industrial cybersecurity. This month, SoftBank group distributor SB C&S announced an agreement to bring Zenity's AI-agent security platform to Japan. On October twenty-first, OMRON announced its venture arm's investment in Israel's Salvador Technologies, which develops solutions for rapidly restoring industrial systems after disruption. That addresses precisely the challenge illustrated by Asahi. These capabilities protect the cloud, factories and the AI agents increasingly acting on its behalf. Let Navon deepen cooperation through joint research, faster warnings and realistic exercises. An attack on one democracy should leave every partner better prepared. Israel already helps protect the free world's cyberspace. Together, Navon can strengthen the services its citizens depend on and the freedom its societies cherish.

More than 800 UK supermarket fridges at heightened risk of hacking. Hackers commandeered the fridge and freezer systems at hundreds of French supermarkets The "front doors" to the control systems of more than 800 UK retailer fridges are "reachable by anyone on the internet who knows where to look", analysis by UK Cyber Defence shared with The Grocer has found. It means the web login page to those fridges' built in computers - known as controllers - which are used by engineers to check temperatures and tweak settings, can be readily found by hackers, rather than being hidden and reachable only via a private, protected connection. The finding comes after hackers commandeered the fridge and freezer systems at hundreds of French supermarkets - including Carrefour, Intermarché, Netto, Coopérative U and Monoprix, according to AFP and Reuters reports - overnight on Sunday, remotely raising their temperature. This led to spoilt food and in some cases fires. It is understood the hackers were able to raise their temperature to as high as 50oC, by forcing the internet-connected fridges into a continuous defrost cycle. Explore related questions. The attacks targeted fridge units manufactured by Danish company Danfoss, and a security flaw in the controllers of some of its models, for which it has issued software updates - known as patches. Present in thousands of stores in France, Danfoss units are less widespread in UK grocery. However, UK Cyber Defence found 15 Danfoss fridges in the UK accessible on the open internet and said it was "accurate to say it is highly unlikely that they have been patched". While having the web login pages discoverable on the internet doesn't automatically give hackers access to the control system of the 800-plus fridges and freezers, it does expose them to "whoever finds the next flaw or guesses a weak password", UK Cyber Defence founder and chief information security officer Peter Bassill told The Grocer. As Bassill put it, it's not a "broken lock" but it is like "having your front door on the street". In a statement to industry publication Cooling Post, Danfoss said the majority of affected units in the France attacks were legacy products or current products running outdated firmware. It is understood hackers found web login pages to the fridges by running scripts on the open internet. Last year, Danfoss issued a software patch to repair a vulnerability that could "result in an authentication bypass". A patch this September was issued to repair a vulnerability that "allows attackers to gain read/write access to internal values, upload and execute unsigned applications". Research presented by cybersecurity company Claroty's Team82 at hacker convention DEF CON 34 in August demonstrated how a compromised Danfoss AK-SM 800A controller could be used by hackers to remotely "configure temperature thresholds" and "adjust operating schedules" as well as "perform maintenance activities". At least 300 Intermarché and Netto brands were targeted in Sunday night's incident, along with more than 100 Coopérative U outlets, and several E.Leclerc stores, the retailers told AFP in separate statements. "Targeting equipment essential to the preservation of fresh products and to the food supply of the French people is extremely serious and is unprecedented. It is unacceptable," Coopérative U CEO Dominique Schelcher said in a statement. Carrefour has said fridges in 80 of its 6,000 stores were hacked, while Monoprix said 5% of its 400-plus stores had been hit, Reuters reports. Attack 'didn't need a genius' According to Bassill, "France was not singled out because someone bore it a grudge; it was singled out because its refrigeration estate was, by a wide margin, the most reachable Danfoss population in the world". UK Cyber Defence's analysis found France alone accounts for roughly 2,967 of the internet-exposed Danfoss controllers worldwide. "The French attack didn't need a genius, it needed a few hundred fridge controllers sitting on the open internet with a known, unpatched flaw, and France happened to have more of those than anywhere in the world," Bassill told The Grocer. "Britain has very few of that particular Danfoss model, but it has over 800 refrigeration and building controllers of one make or another reachable from the public internet, usually put there by a maintenance contractor so they can dial in from the van. The ingredients are the same here; only the brand on the box is different," he added. The majority of the 800 internet-exposed controllers - which cover manufacturers including Danfoss, Carel, Copeland and others - can be found on "consumer broadband, rural broadband and machine-to-machine cellular networks, not on corporate retail infrastructure" UK Cyber Defence said, indicating it is "the work of a refrigeration or building-services contractor who needed remote access and took the quickest route to it". "The retailer's IT team frequently does not know the device exists," Bassill said. "The reassuring part is that this is housekeeping, not rocket science: know which controllers you have, make sure they're patched, and take their management page off the open internet," Bassill added. "Do that and you've shut the door the French attackers walked through." Latest. Partner features. No comments yet. You will be able to: * Read more stories * Receive daily newsletters * Comment on stories

Elisity joins Open Secure AI Alliance to help advance open source tools and standards for AI safety and cyber defense. Company brings a collaborative ecosystem approach to industry coalition SAN JOSE, Calif., Oct. 1, 2026 /PRNewswire/ - Elisity, the identity-based microsegmentation company, today announced it has joined the Open Secure AI Alliance. The Alliance, led by the Linux Foundation, brings organizations together to develop and share open tools, techniques, and technologies that safeguard software and AI agents. Elisity will contribute its perspective on connecting identity and risk context with least-privilege policies, containing agentic threats, and addressing unsanctioned AI use across diverse technology environments. AI agents can act across applications, infrastructure, and services from many providers. Securing that activity requires identity systems, threat detection, application permissions, and segmentation controls to work together. The Alliance supports collaboration across the model, harness, and runtime, as well as the broader software and security ecosystem. "Enterprises need a consistent way to secure AI agents as they move across technologies from different providers," said James Winebrenner, CEO of Elisity. "We have built Elisity around collaboration: bringing together the context our customers' security tools provide and turning it into enforceable segmentation policy. We're joining the Open Secure AI Alliance to bring that experience to the development of open tools and standards for agent identity, least privilege, and containment." Bringing ecosystem experience to a shared challenge Elisity's collaborative approach is reflected in its existing integrations with platforms including CrowdStrike Falcon, Microsoft Active Directory, Armis, Claroty, and ServiceNow CMDB. IdentityGraph correlates identity, asset, and risk context from these complementary sources to inform segmentation policies enforced through supported network infrastructure. This experience informs Elisity's perspective on the information security systems need to exchange so that enterprises can make consistent decisions about what AI agents should be allowed to reach. Through the Alliance, Elisity intends to work with fellow participants on open approaches to three enterprise security priorities: * Least privilege for sanctioned agents. Establish common ways to describe an agent's identity, owner, and authorized scope so that security controls can apply appropriate limits across platforms. Approved agents should be treated as internal actors whose access must be justified by their work and reassessed as that work or its risk changes. * Containment of agentic threats and attacks. Develop shared approaches to communicating risk and coordinating containment across security systems. Enforceable boundaries and interoperable controls are essential to limiting lateral movement and the reach of threats that operate at machine speed. * Identification and control of unsanctioned AI. Identify improvements to how systems share evidence of AI use outside approved deployments, including employee-installed agents and newly connected AI services. Common approaches should help organizations distinguish observed activity from indicators of possible use and connect that evidence to appropriate policy decisions. "Healthcare organizations need to know who owns an agent, what access its work requires, and how to stop it if something goes wrong," said Jason Elrod, CISO at MultiCare Health System and executive adviser to Elisity. "Those requirements span clinical applications, infrastructure, and security tools. Shared standards can help us apply them consistently as AI becomes part of patient care and hospital operations." "A new AI tool can enter through an employee's laptop or a service the organization already uses," said Mike Searight, former CIO of the City of Waco, Texas, and an adviser to Elisity. "Public-sector teams need their technology providers to work together so they can understand that access and limit it. Open approaches give us a better chance of doing that with the infrastructure and staff we already have." Elisity sees participation in the Alliance as an extension of its commitment to helping enterprises put their existing security investments to work together. The company's vision for AI-agent security depends on collaboration across the ecosystem, with shared standards connecting identity, permissions, and enforceable boundaries as adoption grows. Learn more about the Open Secure AI Alliance and Elisity. Frequently Asked Questions About Elisity and the Open Secure AI Alliance What is the Open Secure AI Alliance? The Open Secure AI Alliance is an industry coalition, led by the Linux Foundation, that brings organizations together to develop and share open tools, techniques, and technologies that safeguard software and AI agents. The Alliance supports collaboration across the model, harness, and runtime, as well as the broader software and security ecosystem. Why did Elisity join the Open Secure AI Alliance? Elisity joined the Open Secure AI Alliance to bring its experience turning the context from customers' existing security tools into enforceable segmentation policy to the development of open tools and standards for agent identity, least privilege, and containment. AI agents act across applications, infrastructure, and services from many providers, so securing them requires identity systems, threat detection, application permissions, and segmentation controls to work together. What will Elisity work on through the Open Secure AI Alliance? Through the Open Secure AI Alliance, Elisity intends to work with fellow participants on open approaches to three enterprise security priorities: least privilege for sanctioned agents, containment of agentic threats and attacks, and identification and control of unsanctioned AI. What does least privilege mean for AI agents? Least privilege for AI agents means describing each agent's identity, owner, and authorized scope in a common way, so security controls can apply appropriate limits across platforms. Elisity's position is that approved agents should be treated as internal actors whose access must be justified by their work and reassessed as that work or its risk changes. What is unsanctioned AI, and why is it hard to control? Unsanctioned AI is AI use outside approved deployments, including employee-installed agents and newly connected AI services. It is hard to control because evidence of that use is spread across many systems. Elisity supports common approaches that help organizations distinguish observed activity from indicators of possible use and connect that evidence to appropriate policy decisions. How does Elisity connect identity and risk context to segmentation policy? Elisity IdentityGraph correlates identity, asset, and risk context from complementary sources, including CrowdStrike Falcon, Microsoft Active Directory, Armis, Claroty, and ServiceNow CMDB, to inform segmentation policies enforced through supported network infrastructure. Elisity enforces least-privilege access without agents, new hardware, or network redesign. About Elisity Elisity helps healthcare systems, manufacturers, and critical infrastructure operators stop lateral movement with identity-based microsegmentation. Using identity and risk context for users, devices, workloads, and AI agents, the platform discovers assets, simulates policies, and enforces least-privilege access through existing network infrastructure - without agents, new hardware, or network redesign. Customers include GSK, Main Line Health, MultiCare Health System, and Shaw Industries. Learn more at www.elisity.com. Media Contact: Danielle Ostrovsky Hi-Touch PR [email protected]