X-Industry

Cybersecurity experts at Sophos X-Ops have uncovered a new social engineering campaign that relies on a tactic dubbed ‘TerminalFix’ to compromise enterprise networks.  After Microsoft successfully disrupted fraudulent certificate-signing services earlier this year, threat actors quickly pivoted to alternative methods to infiltrate corporate environments. The latest research details how attackers now deceive users into manually pasting malicious commands into Windows Terminal, triggering a comple

A new AI agent developed to run a full cyber-attack without human intervention has been uncovered by security researchers.  The agent was designed to use hijacked AI models to autonomously run up to 100,000 malicious actions without human approval, including scanning for targets, exploiting vulnerabilities, stealing credentials, deploying cryptocurrency miners and reinfecting compromised systems.[1]

Researchers from the Symantec Threat Hunter Team found the malicious agent using an AI honeypot –

A major cyber-attack on commercial refrigeration systems across France has left more than 150 supermarkets facing millions of euros in damages after hackers remotely changed fridge temperatures overnight. The incident affected cold-chain control software at branches of Intermarché, E.Leclerc, and Coopérative U, also known as Super U. Instead of switching systems off, the attackers increased temperatures inside refrigerators and freezers at hundreds of sites to as high as 50 degrees Celsius overn

The leading identity protection firm Semperis has released its 2026 Ransomware Risk Report exposing how cybercriminals have adopted a new approach to pressure victims into paying ransom. Rather than limiting actions to data encryption or theft, attackers now direct psychological and physical threats towards executives, employees and their families. The study, which surveyed only organizations hit by ransomware in the past year across the US, UK, France, Germany, Italy, Spain, Singapore, Canada a

Yesterday, 21 October, the Internet was having a bad day, from a cybersecurity perspective. Government websites around the world went offline, Reuters reported, after AI discovered critical security vulnerabilities that impacted the programming language PHP.  The issues were severe enough that the US National Security Agency (NSA) advised organizations using PHP on their systems to either immediately update or take their systems offline until they could.[1]

But the PHP update wasn’t the only hea

Most enterprises didn’t choose multi-cloud as a deliberate strategy.  They ended up there: an acquisition brought in a company already running on a different provider, one team picked Amazon Web Services (AWS) because that’s what they knew, another spun up projects in Microsoft Azure because it bundled with existing licensing, and a third quietly built things in Google Cloud Platform (GCP) because a developer liked the tooling.[1]  Nobody sat down and designed this. And nobody ended up owning th

Dutch researchers cited by Reuters report that thousands of wind and solar power systems across Europe are exposed online. The findings have raised questions about how operators can protect increasingly connected energy infrastructure without disrupting the remote access needed to run it.  Wind and solar parks are often unmanned and spread across wide areas. Grid operators, energy traders and maintenance contractors rely on remote connections to monitor equipment and manage operations.  That con

Someone vibe-coded a gun and 3D-printed it and wants the freedom loving citizens of the world to enjoy this power, too.  The creator, gun-rights activist Cody Wilson, told The Verge that it’s the “first AI-made gun” his timely follow-up to the world’s first 3D-printed firearm that he crafted 13 years ago.  Called the WASP, the .22 caliber weapon’s unconventional design aptly looks like it might’ve been hallucinated by an old image model, except it apparently actually works and can fire real ammu

The unnamed US executive is separate from Canadian ransomware negotiator Edward Dubrovsky, who was arrested on federal extortion charges.  US federal investigators suspect executives at two ransomware negotiation firms maintained unusually close relationships with ShinyHunters, including one executive who allegedly helped operate the group’s victim-extortion website.

The findings were reported by Brian Krebs of KrebsOnSecurity today, citing sources close to the federal investigation. The report

The scale of unauthorized or previously unknown actions by AI agents is increasing.   Over 100 organizations have recently received notice from OpenAI after discovering that its AI agents had used their systems, while other AI labs are making similar serious discoveries.  Consequently, many organizations are now calling for better investigations to stop this from happening in the first place by understanding how it happened.  But any investigations are pointless if the AI can’t be tracked.  And

Imagine a lock whose secret is the product of two enormous prime numbers. Anyone can see the product.  Almost no one can recover the original primes. For nearly fifty years, that picture has been the public story of RSA, the public-key scheme that still signs certificates, tokens, and software across the internet.  In their paper “Forging 1024-bit RSA signatures in nearly SNFS time”, researchers Laura Shea, Miro Haller, Adam Suhl, and Nadia Heninger of the University of California San Diego and

Cryptocurrency wallet manufacturer Ledger has confirmed that one impacted device distributed by CryptoBilis, its Southeast Asian reseller, was compromised with an unauthorized hardware implant potentially designed to drain user funds. After receiving multiple user reports of cryptocurrency theft, Ledger has launched a forensic investigation that identified more cases of hardware modifications in devices sold across regional third-party channels.[1]

Update from Ledger Donjon on the CryptoBilis in

Italian Prime Minister Giorgia Meloni is the first major politician to apply to trademark her voice to protect herself against Artificial Intelligence (AI) deepfakes. She has filed an application with the European Union Intellectual Property Office (EUIPO), and it is listed on their website. The prime minister's office has confirmed the application. Meloni recorded a four-second audio in which, in Italian, she says "Io sono Giorgia”, I am Giorgia Meloni, twice.

Meloni is not the first to protect

Hundreds of French supermarkets have been hit by a cyber-attack that caused their fridges and freezers to heat up, spoiling fresh and frozen food estimated to be worth millions of euros.  Hackers gained control of the software that regulates the temperatures of the refrigeration units, forcing some to warm to 50C (120F) overnight, according to one supermarket chain affected.  The chief executive of Coopérative U said several of its supermarkets had been hit by the "extremely serious and unpreced

The Democratic People's Republic of Korea (North Korea) continues to improve its tactics to infiltrate western companies with thousands of state-controlled remote "IT Workers" (ITWs) while avoiding detection.  ITWs are increasingly using facilitators to counter improved private sector detections and law enforcement actions.  The FBI has observed North Korean ITWs use stolen identities, extort victims, and use threats of physical violence to demand payments.  ITWs fund North Korea's government, i

Ontinue’s 1H 2026 threat report shows attackers increasingly abusing Microsoft Teams, npm packages, device-code logins, and blockchain ClickFix tactics to steal access and spread malware.[1]

Cybercriminals increasingly abuse legitimate business tools to access organizations, steal credentials, and spread malware, rather than relying only on software vulnerabilities, according to Ontinue’s 1H 2026 Threat Intelligence Report.[2]  The report describes how these tactics are showing up across workpla

Britain's domestic spy agency, MI5, has issued an unprecedented public warning to British universities regarding a sophisticated espionage operation orchestrated by China's intelligence service.  MI5 has formally identified the China General Technology Research Institute (CGTRI), also known as the China Academy of General Technology (CAGT), as a front organization for the Chinese Ministry of State Security (MSS). The warning represents a rare public attribution by MI5 and underscores the severit

Philadelphia police said a "false homicide tip" submitted through the PhillyUnsolvedMurders.com website was generated by an Anthropic artificial intelligence model.  The AI giant later confirmed the incident in a report, published last week. The AI submission came through an online platform launched by the Philadelphia Police Department to gather tips about open homicide cases.

Anthropic notified the police department of the incident on 7 October, Philadelphia Police spokesman Sgt. Eric Gripp, s

Last week, the DHS Cybersecurity and Infrastructure Security Agency (CISA), Federal Bureau of Investigation (FBI), National Security Agency (NSA), and international partners issued a joint Cybersecurity Advisory warning[1] that Integrity Technology Group (Integrity Tech), a China-based cybersecurity company, is enabling threat actors to target multiple critical infrastructure sectors worldwide using a range of sophisticated tools, such as large-scale botnets, virtual private network infrastructu

Cyberattacks are rising, but the teams defending against them aren't growing fast enough, pushing cybersecurity professionals to the breaking point, according to new research from the Information Systems Security Association (ISACA).   ISACA's State of Privacy 2026 report found that professionals feel increasingly strained, with 65% saying their roles are far more stressful now than five years ago.  Drawing on insights from more than 1,800 privacy professionals across the ISACA community worldwi