New critical vulnerabilities exploited ITW, attacks targeting Italy, APT campaigns and cyberspionage
Vulnerabilities: flaws exploited ITW and new critical issues in PHP, Ruby, NGINX and Java
On 17 and 18 November 2026, CISA added five vulnerabilities exploited ITW to its KEV catalog: CVE-2026-115731 in Perforce Puppet Enterprise, a Deserialization of Untrusted Data vulnerability allowing code execution on Puppet Server through a valid agent certificate, potentially even without authentication in configurations with unsecured autosigning; CVE-2024-10441 in Synology DSM and BeeStation OS, which enables RCE and is included in the advisory referenced by the manufacturer following recent ransomware attacks, without confirmation that this specific vulnerability was exploited; CVE-2026-40217 in BerriAI LiteLLM, which allows attackers to bypass the Python sandbox and execute code, exploited against a honeypot to deploy the XMRig miner; CVE-2026-90436 in Apache Axis 1.x, an Unsafe Reflection vulnerability that may enable unauthenticated RCE through malicious SOAP messages, observed against Atlassian Confluence Data Center and also reported on N-able N-central in non-default configurations; and CVE-2026-90817 in Vanderbilt University REDCap, a Code Injection vulnerability exploitable without authentication by knowing a valid hash of a public survey, also used in an XMRig distribution campaign. In parallel, on 17 November, Java4Shell (CVE-2026-114938) was disclosed, a Heap-based Buffer Overflow in OpenJDK’s Base64 decoder affecting specific versions of the HotSpot JVM on x86-64 processors supporting AVX-512 VBMI. The flaw may enable RCE even without authentication, as demonstrated by researchers on products such as Keycloak, Jira, Confluence and Jenkins. As of 19 November, GreyNoise had detected 187 requests consistent with exploitation attempts originating from five malicious IP addresses, without confirmed compromises. Cloudflare also observed activity associated with Java4Shell and the ARTEX AI framework originating from the same IP address, without demonstrating direct use of the framework to exploit the flaw. On 18 November, the Ruby project also patched seven vulnerabilities, collectively named November2Shell, affecting regular expression and string handling. These included Ruby2Shell (CVE-2026-132825), a one-byte Heap Buffer Overflow potentially exploitable for RCE. RCE chains were demonstrated for five of the seven flaws, but no complete RCE exploits have been publicly released and no ITW attacks have been confirmed. Also on 18 November, five critical vulnerabilities in NGINX Open Source and NGINX Plus, named NGINX2Root and all rated CVSS 9.2, were disclosed: CVE-2026-123007 and CVE-2026-123633 in HTTP/2, CVE-2026-121810 in HTTP/3/QUIC, CVE-2026-123634 in the DNS resolver and CVE-2026-113425 in the mail proxy. The flaws may enable unauthenticated RCE with NGINX worker privileges in specific configurations. Researchers also demonstrated privilege escalation to root and escape from Docker containers by chaining an RCE with a separate 0-day in the Linux kernel’s SCTP subsystem. Public PoCs are available, while on 19 November WASOC reported indications of ITW exploitation of at least one of the HTTP/2 vulnerabilities, without identifying it or providing public technical evidence. Finally, on 19 November, the PHP Security Team patched 18 vulnerabilities, seven of which were critical. In particular, for CVE-2026-126714, named phpAfterShell, researchers documented an RCE chain combining session file manipulation with an additional memory corruption vulnerability in a native PHP class. A public PoC is available for session manipulation, while the complete RCE exploit has not been disclosed. Additional flaws include CVE-2026-129291 (CVSS 8.8), which may enable local privilege escalation to root in PHP-FPM. No ITW exploitation of the new PHP vulnerabilities has been confirmed.
Italy: attacks from multiple threat actors detected
Over the past week, the hacktivist group Dark Storm Team claimed DDoS attacks against Fedespedi – the National Federation of International Freight Forwarding Companies; Auta Marocchi S.p.A. (Autamarocchi); and Autotrasporti e Spedizioni Barone S.p.A.; while the pro-Russian hacktivist collective NoName057(16) claimed two alleged compromises of building automation systems in Italy as part of #OpItaly. Specifically, on 15 November, the group claimed to have obtained full access to the home automation system of a residence allegedly associated with the Italian elite, stating that it could modify air conditioning, swimming pool heating and outdoor lighting. On 19 November, it claimed access to the BMS/SCADA system of an Italian indoor sports arena, asserting that it could interfere with heating, ventilation, the boiler plant and hot water production. Both claims were accompanied by demonstration videos and messages opposing Italy’s support for Ukraine, but there is no independent confirmation of the compromises or actual alterations to the systems. NoName057(16) was also among the entities targeted by new sanctions adopted on 16 November by the Council of the European Union against 30 individuals and two entities under the regimes addressing Russian destabilizing activities and cyberattacks. The collective was sanctioned for its DDoS activities against European institutions, critical infrastructure and other targets. The designated entities also included the Social Design Agency (SDA), considered responsible for disinformation and political interference campaigns directed by Moscow. The measures include asset freezes, financial restrictions and EU entry bans for designated individuals. In response, NoName057(16) described the decision as an “official challenge”, expressing its intention to continue its offensive operations. Turning to the ransomware landscape, n0n claimed on its leak site to have compromised Chiesi Farmaceutici S.p.A. (a possible connection is hypothesized with the claim of 9 November 2026 concerning an unidentified victim); Qilin Team claimed Arrigoni S.p.A.; Aur0ra claimed Sice Previt S.p.A.; while INC RANSOM Team claimed the alleged compromise of Eni S.p.A. In addition, on 17 November 2026, a user identified by the pseudonym flex22 published a post on an underground forum claiming to possess a database containing personal information associated with Poste Italiane. The following day, the same user published another post concerning the alleged disclosure of contracts associated with the company, but the data exposure remains unverified. Also in the ransomware landscape, security researchers observed a campaign in which AuditTeam uses a framework named OpsAgent to compromise Windows systems, encrypt files and present the ransom as a fee for auditing and consulting services. The infections affected organizations across various sectors in Italy, Russia, Finland, Senegal, China, Hong Kong and the Philippines. On its leak site, the group also claimed victims in Turkey, Thailand, Ukraine, Argentina, South Korea, Japan and Spain. Other campaigns targeting Italy included the distribution of the Android malware Mirax, which uses fake Italian Revenue Agency websites and IO app update screens to trick users into downloading malicious APK files. The malware, equipped with banking trojan and RAT capabilities, enables banking credential theft, interception of SMS messages and OTPs, and remote control of compromised devices. In addition, a phishing campaign targeting Italian healthcare professionals through a counterfeit Sistema Tessera Sanitaria website was identified. The operation exploits references to the Ministry of Economy and Finance (MEF) to steal SPID credentials, OTP codes and personal documents. The Conad brand and miPREMIO loyalty program were used as lures in a campaign distributing a new variant of the Android RAT Koktevrat. The threat, already tracked in September 2026 by the Polish CERT, requests Accessibility services that allow it to intercept text, capture screenshots and collect data entered into forms overlaid on legitimate applications, as well as obtain information about the device and installed applications. Finally, Unicoop Firenze, an Italian consumer cooperative operating in the large-scale retail sector, primarily in Tuscany, detected unauthorized access to its Member Management Information System. Investigations confirmed access to certain members’ personal data, including first name, surname, address, email address and telephone number.
APT and cyberspionage: Chinese, Russian and North Korean operations and the use of AI
The Chinese APT Earth Alux conducted a cyberspionage campaign against government entities in Southeast Asia, active since at least July 2025, exploiting vulnerable Microsoft Exchange servers and web servers, as well as phishing techniques, to deploy web shells and the custom RATs PerfectPioneer and KUKURAT. Activities included credential theft, lateral movement, modification of Domain Controller logon scripts and attempts to exfiltrate government documents. Also in the Chinese threat landscape, a recent investigation documented operations against telecommunications networks associated with LIMINAL PANDA and a separate adversary named SilentAir. The former used the Linux backdoor GTPDoor, introducing a variant with C2 communications via GTP-U, while SilentAir exploited mobile operator nodes in at least six Asian countries to conduct interceptions through SS7 and Diameter, primarily targeting Taiwanese subscribers through SMS and call redirection and the acquisition of 2FA codes. Remaining in China, Mustang Panda conducted a campaign likely targeting Philippine government officials involved in preparations for the 49th ASEAN Summit in Manila, using a diplomatic lure impersonating Laos’ Ministry of Foreign Affairs to distribute PlugX through DLL Side-Loading. Also in the Philippines, the evolution of EggStreme, a multistage framework associated with a suspected Chinese APT, was observed. Turning to North Korea, on 12 November the US Department of Justice announced the seizure of cloud infrastructure used in the FaceStealer campaign, attributed with high confidence to Lazarus Group and associated with the theft of more than $400 million in cryptocurrency over the previous 18 months. The operation exploited compromised accounts on messaging and social media platforms to arrange fake Zoom, Teams and Meet meetings, using authentic footage of previous victims and deepfakes to induce targets to install malware. On the Russian front, Callisto compromised two US organizations between late 2025 and 2026, exploiting stolen credentials, MFA fatigue and spear phishing to access Google Workspace and Microsoft Entra ID environments. In the first intrusion, the group obtained Super Admin privileges, exfiltrated documents and established persistence through unauthorized passkeys and MFA devices; it subsequently impersonated an employee of the first victim to compromise a second organization and steal cloud data. In Europe, an investigation into an exposed server in Latvia also documented more than 14 months of activity by the unidentified adversary DarkLunnen, primarily targeting Ukrainian organizations. The attacker exploited CVE-2024-55591 in Fortinet to obtain privileged access to FortiGate firewalls, conducted VNC brute-force attacks and attempted to steal source code from defence and aviation organizations, without evidence of manipulation of industrial processes. Linguistic indicators suggest the use of Russian but do not support definitive attribution. Finally, on 13 November 2026, the FBI reported the malicious use of ARTEX, an AI-based multi-agent penetration testing framework, in cybercriminal and cyberspionage campaigns against organizations in various countries. The platform enables the automation of attack phases, from reconnaissance to data exfiltration, and has been associated with the activities of several groups, including Storm-0940, Storm-0940 and VAPOR PANDA. ARTEX was reportedly also used in a cyber campaign against the European Commission, confirmed on 19 November 2026, which resulted in the temporary shutdown of several internal applications and prolonged service disruptions. The framework’s involvement, reported by the press, has not been officially confirmed, while the full extent of the incidents remains to be established.
Weekly Threats Report is Telsy‘s weekly update covering the main developments in cyberattacks and threat actors worldwide, produced by our Threat Intelligence & Response team.
The team consists of security analysts and researchers with technical and investigative expertise and internationally recognized experience.
Through the continuous monitoring of cyber threats and geopolitical events, it produces and provides organizations with actionable intelligence to anticipate attacks and understand their impact, while relying on a trusted partner in the event of a cyber incident.
Learn more about our Cyber Threat Intelligence solution.
