Researchers warn that hackers could soon exploit a maximum-severity vulnerability (CVE-2026-103849) affecting Magento and Adobe Commerce, potentially allowing them to execute arbitrary PHP code on store servers. [...]

Candiru spyware was delivered to devices at Al Jazeera Media Network via targeted online advertisements in a zero-click attack that required only rendering the malicious ads on the victim's device. [...]

The October 2026 Windows security update may break networking for some customers who are still clinging to long-deprecated NTLMv1 security. A seemingly minor change in the update could have major implications for administrators and enterprises that haven't fully moved away from the legacy technology. Before the October 2026 update, Windows permitted NTLMv1-derived credentials for Single Sign-On and generated warning events. After the update, requests to use NTLMv1-derived credentials are blocked. It comes down to the BlockNtlmSSO registry value. If it hasn't been set, Windows will now use Enforce mode (1) instead of Audit mode (0), thus blocking requests and potentially impacting Wi-Fi, Ethernet, and VPN authentication, and therefore connections. The update does not set the value of BlockNtlmSSO, and administrators can set it to 0 to temporarily allow NTLMv1-derived Single Sign-On credentials. However, Microsoft noted, "Using Audit mode reduces the protection provided by Enforce mode and should only be considered as a scoped, temporary security exception." To be clear, this is not a bug and is by design. The addition of NTLM to the deprecation list occurred in 2024, and Microsoft added audit logs for NTLMv1 usage in late August 2025. The Windows giant has made no secret of its desire to ditch the legacy technology, stating in 2023, "Our end goal is eliminating the need to use NTLM at all." The problem is NTLM's longevity. It appeared in 1993 with Windows NT 3.1, and removing it from enterprise environments has proven tricky, even though it is somewhat insecure by modern standards. Microsoft removed it from Windows 11 24H2 and Windows Server 2025, but some remnants can still be found, notably when using MS-CHAPv2-based connections, including PEAP-MSCHAPv2, and EAP-MSCHAPv2, in a domain-joined environment. Hence the introduction of the audit mode in 2025, to help administrators spot where the technology was still being used. It is now 2026, and Microsoft has taken the next step. If Credential Guard is not enabled and BlockNtlmSSO is not set to Audit mode (0), the use of NTLMv1-derived credentials will be blocked as Microsoft works "to finalize the disablement of NTLMv1." There is no indication of how long the BlockNtlmSSO workaround (setting the value to Audit mode) will work - most likely until the next major Windows release. In September, Microsoft said, "In-market versions of Windows client and Windows Server will continue to support NTLM." However, it also stated, "Our long-term objective is a Windows platform that doesn't need NTLM at all." As such, if the message wasn't already clear, the move to enforce it by default makes it impossible to miss. The technology still exists in the operating system, but it is on its way out and will be removed in a future Windows release. Problems with networking after the October 2026 Windows security update could therefore signal that migration plans are not as far along, or as complete, as administrators might have thought. ®

Employees using personal ChatGPT accounts at work can expose sensitive company data, even when their organization provides an approved Enterprise workspace. NordLayer explains how IT teams can investigate unauthorized AI use and block sensitive information before it reaches personal accounts. [...]

The US' bankruptcy courts are currently down owing to what officials are calling a “cybersecurity incident.” The Administrative Office of the US Courts said the websites and systems that support the Bankruptcy Noticing Center (BNC), which manages electronic comms related to bankruptcy cases, were pulled offline while the vendor investigates. BAE Systems runs the affected systems and websites on behalf of the Federal Judiciary, and is “working to restore operations,” according to an announcement by the US Courts on Thursday. “As a result of the incident, all access to the BNC systems and websites that BAE operates on behalf of the judiciary and the related bankruptcy noticing process are unavailable at this time, including bnc.uscourts.gov, bankruptcynotices.uscourts.gov, notices.uscourts.gov, and the National Creditor Registration Service," it said. “The incident has also halted the BNC’s production and distribution of bankruptcy notices for bankruptcy courts.” While the public notice was published on Thursday, the issues affecting the BNC systems began a few days earlier, on October 9, according to the Bankruptcy Court of the Eastern District of Texas. Alongside its own announcement, the same court published a general order indicating that for any case filed on or after September 14 with an active motion to stay, that delay would be extended by an additional 30 days. The court said it will continue to accept filings, open cases, and process documents, but said there may be a delay in the entry and service of notices and orders until the issues are resolved. The courts' ability to process electronic and paper bankruptcy notices is affected, and it applies to federal bankruptcy courts across the country. A spokesperson at BAE told The Register: "BAE Systems, Inc. operates the Bankruptcy Noticing Center (BNC) on behalf of the Administrative Office of the U.S. Courts (AOUSC). We are aware of a security incident that is isolated to the BNC system and have taken immediate steps to address it. We are coordinating closely with AOUSC on the response." The disruption comes at a difficult time for the US public sector, which has struggled with multiple recent cyberattacks. ShinyHunters' attack on the FBI in September is doubtless the most high-profile of these, although the Arizona Judicial Branch confirmed an attack on the state’s court systems at around the same time. Officials said the “criminal hackers or their bots” were responsible for stealing personal data and backup court files. The US Courts also said in August 2025 that it was stepping up its cybersecurity measures in response to increased targeting of its case management system. ®

Burger King Japan, the Japanese subsidiary of global fast-food giant Burger King, disclosed a major data breach after attackers accessed a server containing 11.18 million customer records. [...]

Meta e OpenAI affidano a personaggi teneri il compito di rendere più familiari gli agenti capaci di gestire posta, agenda e pagamenti. La strategia cerca di superare la diffidenza degli utenti, ma apre interrogativi su dati, minori e controlli effettivi

L'articolo Mascotte per gli agenti AI: un volto tenero dietro un potere crescente proviene da Agenda Digitale.

L'intelligenza artificiale in Umbria entra nella fase delle applicazioni concrete. L'esperienza maturata dalle PMI con Industria 4.0 mostra che tecnologia, processi, competenze e reti territoriali devono avanzare insieme perché l'innovazione produca valore

L'articolo Dalla fabbrica connessa all’AI, il percorso di innovazione delle PMI umbre proviene da Agenda Digitale.

An old-fashioned smash-and-grab job on a photographer’s vehicle threatens the security of pupils’ data as a UK secondary school says it faces a potential data protection pickle. Jon Curtis-Brignell, headteacher at Priory School in Lewes, said in an email this week that a company the school hired to take pupils’ annual school photos has alerted it to a data theft case after a vehicle was broken into. “Pret-a-Portrait has informed us that, overnight between 6 and 7 October, one of its vehicles was broken into and a camera bag containing SD memory cards was stolen,” Curtis-Brignell wrote. “Some of these cards may have contained photographs taken at Priory School, together with images of the QR code cards used for school photographs, which show pupils’ names and class names.” The headteacher said that the same equipment used to take the photos at Priory School was also used for three additional photography jobs after Pret-a-Portrait visited the school on October 1. Pret-a-Portrait, which has offices in the UK, Qatar, and the UAE, told the school that it believes “most, if not all” of the files containing pupils’ data will have been overwritten as a result, but it cannot be certain. Curtis-Brignell said that the potentially affected information included photographs, pupils’ names, and their “class information associated with the photography process.” Other types of data, such as addresses, contact details, financial information, medical information, and safeguarding records stored in the school’s own systems remain untouched. Nick Kerr, owner of Pret-a-Portrait, told The Register that no physical, printed photos were stored in the vehicle at the time of the break-in, and that it was not possible to determine the potential scope of the data protection violation. “As the email from Priory School explains, it is likely to be somewhere between a small subset of their pictures and none of them,” said Kerr. On a positive note, he said that there is nothing to suggest the files have been improperly accessed. “Our analysis reveals that no improper access to pictures from this job occurred in between the time of the theft and the pictures being taken offline.” Per the headteacher’s email, Pret-a-Portrait confirmed that it took a number of steps to mitigate the potential damage, including taking all school photos offline, changing the IDs used to access them, and changing its procedures so that all SD cards are wiped at the end of each commission. The company also “reviewed and strengthened its procedures following the incident,” and reported the matter to both the police and the ICO. “I appreciate that receiving an email of this nature may understandably cause concern,” Curtis-Brignell added. “We have worked closely with our Data Protection Officer and Pret-a-Portrait to understand what happened, establish the potential risk and ensure that appropriate action has been taken. “We wanted to be open with families about what has happened, while also reassuring you that there is currently no evidence that the information has been accessed or misused and that a number of steps have now been taken to prevent access to the affected photographs and reduce the risk of a similar incident occurring again.” ®

Le API key di accesso ai modelli di AI rubate stanno diventando un obiettivo primario per i cybercriminali. Secondo una ricerca di Bitsight, tra febbraio e agosto 2026 i forum underground hanno registrato un aumento del 67% di discussioni su agenti AI, tecniche di prompt injection indiretti e protocolli come il Model Context Protocol (MCP).

Ma quali rischi comporta questa tendenza e come possiamo proteggere i nostri sistemi?

Uno strumento che trasforma le API key rubate dall'AI in minacce automatizzate

In alcuni forum riservati viene offerto Monst3r-AI, un tool di penetration testing che opera in modo completamente autonomo. Gli attaccanti inseriscono API key rubate dall'AI in un pannello di controllo, impostano un budget massimo e indirizzano lo strumento verso i target prescelti. Il vantaggio per i criminali? Tutti i costi di calcolo ricadono sulla vittima, bypassando i controlli di sicurezza di provider come OpenAI, Anthropic e Google.

Il venditore dichiara di aver ottenuto 1.072 risultati su 199 host scansionati, ma Bitsight non può confermare questi dati. Tuttavia, anche una minima parte di queste informazioni potrebbe essere sufficiente per rendere il tool estremamente pericoloso. Gli esperti prevedono che le API key rubate tramite AI verranno sfruttate sempre più spesso per alimentare scanner simili, trasformando le vittime ignare in "bancomat" involontari.

Jailbreak e modelli open sono le nuove frontiere del cybercrimine

Le discussioni sui jailbreak, tecniche per bypassare i limiti di sicurezza dei modelli di intelligenza artificiale, restano le più popolari sui forum underground con un aumento del 75% nel periodo analizzato. Tuttavia, la tendenza emergente riguarda i modelli open-weights come Glm e DeepSeek, sempre più richiesti per la loro flessibilità.

Su board cinesi e russe, gli attaccanti più esperti stanno già valutando i modelli in base alla loro "docilità" verso richieste pericolose. Alcuni arrivano a rimuovere manualmente i comportamenti di sicurezza, una pratica nota come abliteration. L’accesso a questi strumenti è poi reso ancora più economico: un rivenditore come CloseRouter offre modelli sia open che commerciali a circa 10 centesimi per milione di token, contro oltre 1 dollaro dei servizi ufficiali.

Tendenze in crescita e in calo di AI e ransomware

Analizzando il vocabolario dei forum criminali, Bitsight ha rilevato una crescita del 45% nelle discussioni sull’intelligenza artificiale, mentre i riferimenti a ransomware sono crollati del 49% e quelli a phishing dell’8,5%. Questo cambiamento potrebbe essere in parte attribuibile alle azioni di contrasto come i takedown, ma suggerisce anche uno spostamento di attenzione verso nuove minacce.

Anche i chatbot criminali come WormGpt e FraudGpt hanno registrato un calo dell’8,7%, probabilmente a causa di incidenti come la violazione di WormGpt all’inizio del 2026, che ha esposto i dati sensibili di circa 19.000 utenti. Tuttavia, la minaccia non è scomparsa: sta evolvendo verso strumenti più sofisticati e difficili da tracciare.

Come proteggersi dalle API key rubate dall'AI

La priorità è trattare le API key dell'AI rubate con la stessa attenzione riservata a una credenziale Aws violata. Monitorare e revocare tempestivamente le chiavi compromesse può prevenire danni incalcolabili. La difesa, tuttavia, non si ferma qui.

Ecco le azioni concrete da intraprendere oggi:

1. Verifica l’origine di ogni server MPC e plugin di IA: Non fidarti ciecamente dei plugin scaricati da repository pubblici. Ogni strumento deve essere sottoposto a una revisione manuale approfondita e, se possibile, testato in un ambiente isolato. La sicurezza non può essere delegata alla fiducia verso comunità open source o marketplace di terze parti.

2. Blocca l’accesso degli agenti locali a directory sensibili: Gli agenti AI che operano in locale rappresentano un rischio elevato se hanno accesso a file system o comandi shell senza supervisione umana. Imposta politiche di blocco automatico per le directory critiche e limita l’esecuzione di comandi pericolosi a operazioni approvate manualmente.

3. Monitora i login per eventuali API key rubate dall'AI: Gli agenti AI che guidano browser headless operano con una cadenza distinta rispetto agli utenti umani. Tenere sotto controllo i tassi di accesso su Notion, Slack e GitHub può aiutare a identificare attività sospette prima che causino danni. Strumenti di anomaly detection basati su comportamenti anomali sono ormai essenziali.

I prossimi passi per proteggerti

La minaccia delle API key di AI rubate non è più un rischio teorico: è una realtà che sta già colpendo aziende e professionisti. La chiave per proteggersi è agire proattivamente, monitorando costantemente, aggiornando i sistemi e formando i team su queste nuove sfide.

Il futuro della sicurezza informatica passa inevitabilmente dall’AI, ma spetta a noi decidere se sarà un’arma o uno scudo.

L'articolo API key rubate dall'AI per attacchi autonomi dei cybercriminali proviene da sicurezza.net.

The Administrative Office of the US Courts says a cyberattack on government contractor BAE Systems has taken down the notice system used by federal bankruptcy courts. [...]

BCE, Banca d’Italia e IVASS chiedono alle banche di rafforzare la resilienza digitale davanti a minacce informatiche accelerate dall’AI. La risposta deve unire valutazione dell’esposizione, priorità di intervento, Risk Appetite, coordinamento tra funzioni e informazioni utili al Consiglio di Amministrazione

L'articolo Rischio AI nelle banche: la resilienza digitale deve cambiare passo proviene da Agenda Digitale.

Change is constant, but The Register likes to keep some things the same – such as using Friday mornings to bring you a new installment of On Call, the reader-contributed column that shares your tech support stories. This week, meet a reader we'll Regomize as "Daniel" who told us that in 2020 he was hired to replace a two-person IT department at a UK school. "The IT manager had left amid some 'accounting irregularities,' mainly involving high-end gaming PCs being purchased using school funds but never making it onto the school site," Daniel told On Call. "His deputy stepped up just in time to have to deal with a ransomware attack that took the school's entire network offline for three months." The school eventually hired external contractors to rebuild the network and restore from the only available backup, which was months old. The deputy IT manager then quit without serving their notice period. Daniel took the vacant job and in his first week the school principal asked him to audit the network and deliver a report explaining how to get things up to scratch. "I found a few quick wins, including the password policy which at the time was set to six letters. Just that, no complexity, no expiry," Daniel told On Call. "And users hadn't been forced to change them since before the ransomware attack. Or, indeed, ever." The principal approved Daniel's proposed password policy and a requirement for everyone to change their password immediately. "I was not prepared for the backlash this would create," Daniel wrote. The school's data manager proudly said she had used the same password since 1998 and justified keeping it on the grounds that everyone on staff knew it. "She said, 'If I'm not in, they can log in as me and run the reports.'" It gets worse: the password was "letmein" and the data manager had written it on a piece of paper and taped it to her desk. As Daniel rolled out the new password policy, more complaints rolled in. The school principal eventually told Daniel to direct all enquiries regarding the password changes to him. "Within two days, he came into my office, sighed, put his head in his hands, and ordered me to revert the change and to remove the requirement for everyone to change their password," Daniel wrote. "I resisted, I really did," he wrote. "But it was his school. I got him to put the request in writing." The principal left three months later, and a nonprofit took over the school. Daniel's new bosses agreed with his stance and reinstated his password policy. The data manager again fought the change. "After a meeting with me and the nonprofit's IT director, she decided to take early retirement, so she wouldn't have to learn the new systems – or change her beloved password." Have management or users rejected your advice? Click here to tell us what happened, and your story could appear in a future edition of On Call. ®

The Thoma Bravo-owned Proofpoint will acquire Varonis in an all-cash transaction and Varonis will become a private company.

The post Proofpoint Inks $6.3 Billion All-Cash Deal to Acquire Varonis appeared first on SecurityWeek.

BeyondTrust warned customers on Thursday to patch a critical flaw in its Remote Support (RS) that has been exploited in zero-day attacks since late August. [...]

Security researchers warned about exploitable vulnerabilities in refrigeration controllers months - and in one case, more than a year - before cyberattacks disrupted hundreds of French supermarkets. Late Sunday night, unknown criminals hacked thousands of controllers, cranking the heat as high as 122°F (50°C), spoiling food and sparking fires across nearly 700 of the country’s supermarkets. France's national cybersecurity agency ANSSI on Monday warned of widespread compromises involving Danfoss refrigeration controllers, and a day later added affected systems from Copeland and its Dixell brand. The compromised devices were accessible over the internet, with many reportedly no longer receiving software or security updates and others running outdated firmware. Danfoss did not immediately respond to The Register’s request for comment. Copeland emailed us the following statement: “We acknowledge the vulnerability identified in a discontinued, end-of-life generation of controllers. Protecting our customers’ operations has always been a top priority. We are currently working with impacted customers to mitigate the risk. We will continue to support customers in strengthening cybersecurity through supported products, current software and industry-recommended practices.” Operational technology (OT) security firm Claroty’s Team82 sounded the alarm on these buggy devices in August, warning that hackers could “freeze the controller” and “defrost the food.” At the time, the researchers published details of dozens of vulnerabilities, including authentication bypass and operating-system command execution flaws, affecting Danfoss AK-SM 800A and Copeland XWEB controllers. Even earlier, the US Cybersecurity and Infrastructure Security Agency (CISA) published advisories about security flaws in both brands’ controllers. In February, CISA said nearly two dozen CVEs in Copeland XWEB and XWEB Pro could be exploited to “allow an attacker to bypass authentication, cause a denial-of-service condition, cause memory corruption, and execute arbitrary code.” In a May 2025 advisory, updated in August 2025, CISA published a similar warning about buggy Danfoss AK-SM 8xxA series devices. “What makes this incident particularly striking is the timeline,” Tiago Henriques, chief underwriting officer at cyber insurance provider Coalition, told The Register. He noted that the French supermarket hacks happened two months after Team82 published its research, and Danfoss released a firmware update (R4.3.1) to mitigate the flaws. 'Critical exposure' continues “Yet, despite the availability of these patches, Coalition’s scanning identified critical exposure to these threats continuing well into the fall,” Henriques said. Between September 7 and October 6, Coalition’s sensors identified 1,506 unique IP addresses still publicly exposing a Danfoss StoreView web-accessible admin panel, and 820 of these - 54 percent - were associated with French networks. “Our scans counted at least 2,624 refrigeration controllers from the named brands with a login page open to the internet: 1,506 Danfoss StoreView and 1,119 XWEB,” he added, noting that one IP runs both types of controllers. To be clear: this doesn’t mean all of these controllers were vulnerable, but rather they were exposed to the open internet. Even after the cyberattacks - as of Thursday - 618 Danfoss devices remain connected to the internet, according to the insurance provider’s count. Of these, 206 are connected to French networks, while 50 are in the US, and two in the UK. “An unpatched device featuring a web-accessible admin panel can act as a major entry point” for digital attackers, Henriques said. “The takeaway is clear: digital risk extends far beyond workstations and email servers.” Criminals are targeting OT and other industrial devices that are connected to the internet, too. “Identifying these exposed access points before they are targeted can directly contribute to preserving business continuity,” Henriques said. OT cyber leaders call for CISA directive In America, a group of OT cybersecurity and technology providers doubled down on their call for stronger food and agriculture protections in the US. The French cyberattacks “are another stark reminder of the importance of implementing baseline cybersecurity measures to enhance operational technology cybersecurity,” the Operational Technology Cybersecurity Coalition (OTCC) said in a Wednesday statement. Earlier this month, the group, whose members include Claroty, Forescout, Honeywell, and Schneider Electric, called on CISA to issue a Binding Operational Directive to improve OT cybersecurity within the federal government. The OTCC repeated this demand on Wednesday and suggested that the US Department of Agriculture could encourage organizations in the food and agriculture sector to implement cybersecurity best practices. “While no incidents have been reported in the United States, malicious actors are undoubtedly viewing this incident as a potential attack vector to perform here,” the statement said. “We cannot wait for that to happen.”®