Back News.Risky.Biz Risky Bulletin: New Windows security feature will allow apps to refuse to run on insecure systems
Microsoft has developed a new feature that will let apps test if certain security features are enabled on a Windows device and then refuse to run, disable some of their own features and capabilities, or ask the user to take action and update their OS settings.
The new feature was announced this week and is named the Windows Baseline Protection .
It is an attestation process where Windows checks what security features are enabled at startup and then provides the result to apps and local admins. According to Microsoft, Baseline Protection currently checks for eight items, but new checks will be added in the future.
Currently, Baseline Protection is capable of testing for Secure Boot, a security processor, signed or Windows-certified drivers, and if memory integrity is on, virtualization-based security is on, and if the kernel debugger and vulnerable driver protections are enabled.
Some of these checks are available in Windows already and are used in various parts of the Security settings panel. The difference to Baseline Protection is that these checks are auditable by third-parties and not Windows itself.
Right now, app developers usually have their own scripts and methods for checking for these security features, which are often buggy and can crash Windows systems.
Microsoft hopes this data will be used by app developers going forward to create more secure apps while keeping the stability of its OS.
Password managers, banking and document e-signing, gaming anti-cheats, and exam proctoring software will be able to check if the features they want are enabled and trust the result.
Baseline Protection is out in preview and will be generally available for all users year.
Risky Business Podcasts
The main Risky Business podcast is now on YouTube with video versions of our recent episodes. Below is our latest weekly show with Pat, James at special guest co-host Jason Kikta the helm!
Breaches, hacks, and security incidents
Breach impacts 50 Irish credit unions: Hackers have stolen sensitive data from almost 50 credit unions in Ireland after breaching a shared IT provider. According to Irish public broadcaster RTE, the breach took place at a company named Wellington IT. Several of the affected credit unions have notified the Irish government and are investigating the hacks. Others learned from the breach from the government. [ RTE // St. Canice's Credit Union ]
CodinGame breach: AI platform CodinGame has notified users of a security breach. An attacker exploited a data analysis tool to access the platform's internal network and copy data from its databases. Stolen data includes user details and their activity on the platform. [h/t chum1ng0]
CodinGame ha estado notificando a sus miembros por un incidente de seguridad #Codingame #databreach — chum1ng0/security research (@chum1ng0) November 15, 2026
CodinGame ha estado notificando a sus miembros por un incidente de seguridad #Codingame #databreach
Chargemap breach: Chargemap, an app for finding charging stations and charging stations, is notifying customers of a breach. [ h/t Renee 420 ]
Ha tiens, @Chargemap c’est fait péter — Arnaud de Bermingham (@a_bermingham) November 18, 2026
Ha tiens, @Chargemap c’est fait péter
N-able gets hacked: Hackers have breached the cloud platform of cybersecurity and IT provider N-able. The attackers accessed Mail Assure and SpamExperts servers in a breach on Monday. N-able says it found no evidence that customer details were stolen. The company is now slowly rebuilding the affected services. [ N-able ]
Equinor cyberattack: Norwegian oil and energy giant Equinor suffered a cybersecurity breach but says the incident has not impacted production capabilities. [ Equinor ]
Ethiopia Airlines hacked: Hackers have breached Ethiopia Airlines' ShebaMiles loyalty programme and stole customer data. [ Capital News ]
KPMG Canada gets hacked: Accounting giant KPMG's Canadian is notifying customers of a security breach. Did not say what happened. [ The Globe and Mail ]
Rakuten gets hacked: Hackers have stolen data on more than 15 million customer accounts from a test environment at Rakuten's Securities division. More than 4.4 million of the accounts contained sensitive information. No passwords, PINs, and other authentication data was stolen. [ Rakuten Securities // The Japan Times ]
BundID hack impacts 10,000: Hackers have compromised at least 10,000 user accounts on Germany's e-government platform BundID. Officials have locked the affected accounts and are notifying owners. The platform was hit by a brute-force attack over the past weekend. [ Heise // Handelsblatt ]
Hackers hit EU HR platform: A cyberattack has taken down the European Commission's internal HR platform. EU officials have not been able to access the Sysper platform for almost a week. The platform is used for managing timesheets, managing salaries, signing up for training, and requesting holiday or sick leaves. The hack took place at the end of October and involved the ARTEX AI tool. [ Euronews ]
NVIDIA getting ransomed: The Gentlemen ransomware cartel has listed American chipmaker NVIDIA on its dark web leak site. The company has not yet confirmed any security breach.
General tech and privacy
Chrome 158: Google has released version 158 of its Chrome browser. See here for security patches and webdev-related changes . The biggest change in this release is the addition of a new HTML element to, obviously, launch installation flows. There's also a new WebHID API to allow websites to interact with Human Interface Devices (HIDs) devices (MnK, gamepads, or touchscreens) over USB and Bluetooth.
Ubuntu to let AI fix patches in Edge releases: Canonical will ship AI-generated security patches in Ubuntu Edge releases. All fixes will eventually be reviewed by a human before they make it in the mainline releases. The company cited the increased number of AI-found bugs and the need to provide temporary fixes for all of them. [ Ubuntu // HelpNetSecurity ]
French company fined €5m for plaintext password storage: France's privacy regulator has fined an unnamed company €5 million for storing passwords in plaintext. [ CNIL ]
AI agents produced terrorist plans: The agents of major AI companies produced detailed terrorist plots when prompted during a test by the US Government Accountability Office. The test included ChatGPT, Claude, Gemini, and Grok. The agents produced the plans within a few hours and included detailed steps to carry out mass casualty attacks. [ US GAO ]
Gemini 4 Neon: Google has launched Gemini 4 Neon, a new frontier AI model that can do, among other things, "fast, low cost, frontier-level cyber defense." [ Google ]
Telegram working on SIM card checks: Instant messaging service Telegram is working on a new feature to verify and link user accounts to a phone's SIM card. The new system is being developed to comply with Indian legislation. The Indian government gave Telegram and other messengers until the end of the year to comply with the new rules and roll out the feature. [ Telegram Info ]
Government, politics, and policy
New US Threat Disruption Center: The US government has launched a new initiative to disrupt and take down scams and sextortion schemes. The new Threat Disruption Center will operate under the US Homeland Security Task Force. It will bring together US law enforcement agencies and the private sector to identify and take down scam networks faster. [ US HSTF ]
LoSapio disastrous hearing: Palantir's former CISO Michael LoSapio had a disastrous confirmation hearing at the Homeland Security and Government Affairs Committee. The White House nominee for CISA Director was deemed "disqualified" after he refused to say who won the 2020 election and if the Capitol was attacked on January 6. [ CyberScoop ]
Russia accuses Wikipedia of trying to destabilize the country: Russia's Prosecutor General has accused Wikipedia of running an information campaign to destabilize the country and that it may be labeled an undesirable organization. [ Interfax ]
In this Risky Business sponsor interview , Catalin Cimpanu talks with Thinkst CTO Marco Slaviero the state of the defensive AI cyber space and the company's current agentic canaries, as well as the need for more experimentation.
Arrests, cybercrime, and threat intel
Russian arrested for spying on US tech firm: US authorities have arrested a Russian national who conducted illegal surveillance of a US tech company and defense contractor. Igor D’iu entered the US illegally through Mexico and received orders from his handlers via Telegram. He bought surveillance equipment and spied on the company's employees near their work places and at their homes. D’iu received orders from Russian intelligence and spied on the target because it was supplying weapons to Ukraine. The targeted company is believed to be Palantir. [ DOJ ]
Scattered Spider member gets six years: A US judge has sentenced a Scottish man to six years in prison for hacking a dozen companies and stealing more than $11.6 million worth of crypto assets from victims across the US. Tyler Robert Buchanan carried out the hacks as a member of the Scattered Spider hacking group. Buchanan and accomplices hacked telcos, tech companies, and cloud providers between 2021 and 2023. They used the access to hack individual customer accounts, bypass 2FA, and hijack their cryptocurrency wallets. He's the fifth Scattered Spider member to be sentenced to prison. [ DOJ ]
Fake bomber detained in Ukraine: Ukraine's cyber police arrested a man who sent bomb threats to thousands of schools in the Czech Republic, Slovakia, and Latvia last year. He used email addresses and Telegram to send the threats. The suspect was initially identified in July of last year but fled to Belarus to avoid authorities. He was arrested in September after returning to Ukraine. [ Ukraine Cyber Police // Eurojust // Anadolu Ajansi ]
BEC scammers detained in NC: US authorities have arrested two suspects linked to BEC scams against US healthcare orgs. [ North Carolina State Bureau of Investigation ]
Violent incidents linked to Revolut hack: Dutch Police have arrested multiple individuals who were planning violent incidents across Amsterdam and The Hague. The suspects were using data stolen from Revolut in September to target high-earning individuals. At the time, hackers posed as law enforcement officials to obtain the personal data of selected Revolut customers. Officials said the suspects are French and Belgian nationals. [ Dutch Police // AD.nl // Reuters on the Revolut breach ]
Ransomware affiliates arrested in Ukraine: Ukrainian authorities have arrested four members of a hacking group linked to ransomware and data extortion attacks. Officials say the group made over $40 million from attacks on 28 companies alone. The group allegedly has 27 members, including former members of the Conti ransomware gang. [ Ukraine Cyber Police // Ukraine National Police ]
Dizigom takedown: Anti-piracy alliance ACE has taken down another illegal streaming portal. This time it was Dizigom, Turkey's largest piracy streaming platform. [ ACE ]
AI toolkit used to hack African stock exchange: Broadcom security researchers have spotted an AI hacking toolkit made up of 41 PowerShell modules that was used to breach an unnamed African stock exchange in just three minutes. [ Broadcom ]
Coinbase blocks Musk scam: Coinbase says it blocked $4.7 million worth of transactions where customers would have sent money to a Musk and Tesla-themed scam. [ Coinbase ]
Microsoft Python package spreads infostealer: A hacker has compromised a Microsoft Python library for building AI agents. The incident took place on Tuesday and impacted the Microsoft Agent Framework PyPI package. The malicious payload was an infostealer that stole cloud and CI/CD credentials. [ OpenSourceMalware // Elastic // Ox Security // SafeDep // Step Security // Socket Security ]
ARTEX now supports PHP2Shell: The ARTEX AI pen-testing framework has been updated with support for PHP2Shell attacks. [ Netskope ]
AuditTeam profile: A new cybercrime group named the AuditTeam has been hacking and extorting companies across the world under the guise of "security audits." The group uses the OpsAgent framework to breach targets and demands payment to release the results of their audit. Companies that don't pay are listed on their dark web leak site and have sensitive data leaked online. The group has been active since January this year and has made tens of victims. [ Kaspersky ]
Optimizely hack linked to Caishen: Security firm Censys has linked the Optimizely supply chain incident to Caishen, a new iOS hacking toolkit that blends both Coruna and DarkSword. [ Censys ]
Malware technical reports
BubbleLoader: ReversingLabs analyzed the latest variant of BubbleLoader, first spotted mid-September. [ ReversingLabs ]
BadVibe: Security firm Kaspersky has found a new malware strain named BadVibe that is being spread through Telegram spam and YouTube videos. [ Kaspersky ]
HDGfx Infector: Researchers have spotted a new file infector and USB worm that appears to have been built from scratch this year. [ OALABS ]
Bazuka: Security researchers have discovered the first LLM-based malware framework for the Android ecosystem. Named Bazuka, the framework comes with its own dropper and 23 modules to cover anything from showing overlays to 2FA bypassing and clipboard hijacking. [ ThreatFabric ]
In this Soap Box edition of the podcast, Patrick Gray chats with Thinkst Canary founder Haroon Meer his "decade of deception."
APTs, cyber-espionage, and info-ops
UAT-12643: Cisco Talos has discovered malicious infrastructure linked with a suspected North Korean financially motivated threat group tracked as UAT-12643. [ Cisco Talos ]
Cold Relic: Google's Mandiant division looks at a Cold Relic (COLDRIVER/UNC4057/Star Blizzard/Callisto/Gossamer Bear) intrusion from last year. [ Google ]
Iranian hacking groups: AttackIQ researchers have published a top-down view of Iran's cyber groups and their operations. [ AttackIQ ]
Vulnerabilities, security research, and bug bounty
Security updates: Adobe , Apache Roller , Cisco , Chrome , Drupal , FreeBSD , Microsoft SharePoint , MOVEit , Nginx , Oracle , PHP , Ruby , Synology , Ubiquiti , VMware .
Apache Roller zero-day: The Apache Software Foundation has released a security update for the Apache Roller blogging platform to fix an actively exploited zero-day. The vulnerability can allow attackers to run malicious code on a blog. It is a classic untrusted deserialization attack. The Apache Roller team says an attacker took a bug it patched in September and modified it to work on more older versions. The zero-day now impacts all Roller versions released over the past two decades. [ Apache Roller // CVE-2026-129448 ]
VMware zero-day exploited by Chinese APT: Broadcom has released a security update for the VMware vCenter Server that was exploited in the wild by a Chinese APT group. The zero-day allowed the group to bypass authentication and access the server's directory service. Google linked the attacks to a group it tracks as Armory Castle. The same group was also behind a zero-day in BeyondTrust remote access applications last month. [ VMware // Google ]
FastUnlock zero-day: Security researcher Nightmare Eclipse has released a new Windows zero-day named FastUnlock. The zero-day can allow attackers to bypass the lockscreen and Bitlocker and execute code on the device. It exploits the Windows Narrator service. [ GitHub ]
NGINX2Root vulnerabilities: A collection of five vulnerabilities can allow attackers to gain root access on NGINX web servers. Named NGINX2Root, the vulnerabilities target the server's HTTP/2 implementation. NGINX2Root works on most servers and only requires that HTTP/2-over-TLS is enabled. A PoC is publicly available on GitHub. NGINX is the world's most widely used open-source web server, running on more than 225 million websites. According to Censys, there are more than 9.4 million NGINX servers that run vulnerable versions. [ NGINX2Root // Rapid7 // Wiz // Xint // F5 patches // Censys ]
Pre-auth RCE in nginx is the bug every hacker dreams of finding. It almost never happens. Today we’re disclosing nginx2root: five nginx vulnerabilities, and an exploit that takes an unauthenticated attacker all the way to root. One of themhits the default config. If you run nginx, patch now. Then read: CVE-2026-123007, CVE-2026-123634, CVE-2026-123633, CVE-2026-121810, CVE-2026-113425 — Brian Pak (@brian_pak) November 18, 2026
Pre-auth RCE in nginx is the bug every hacker dreams of finding. It almost never happens. Today we’re disclosing nginx2root: five nginx vulnerabilities, and an exploit that takes an unauthenticated attacker all the way to root. One of themhits the default config. If you run nginx, patch now. Then read: CVE-2026-123007, CVE-2026-123634, CVE-2026-123633, CVE-2026-121810, CVE-2026-113425
ruby2ßhell vulnerability: The Ruby programming language has fixed a memory corruption bug that can let attackers run remote code on Ruby apps. Named ruby2Shell, the vulnerability impacts many Ruby apps and frameworks in their default configurations. Confirmed exploitable include Ruby on Rails, Sinatra, GitHub Enterprise, and GitLab. [ Veria Labs // Ruby security advisory ]
phpAfterShell vulnerability: Security firm Depthfirst has disclosed a new remote code execution bug in the PHP programming language. Named phpAfterShell, the vulnerability can be used against servers that use PHP file-based sessions. The PHP project released patches for phpAfterShell and 16 other vulnerabilities on Thursday. [ Depthfirst ]
Sign2Shell: Security researchers have found RCE vulnerabilities, all rated 9.6/10, in the five most-installed digital signing extensions and apps in Brazil. This includes Lacuna Web PKI, Softplan Web Signer, Certisign WebSigner, Certisign CertiPlugin, and BRy Assinatura Digital. [ Am I Being Pwned? ]
Drupal CMS patches RCE: The Drupal CMS has released a security update to patch a vulnerability that can enable remote code execution attacks on Drupal sites. The vulnerability allows threat actors to abuse the Drupal installer to reinstall the site and run their own code. The bug is not universal and only exploitable in certain scenarios. [ Drupal CVE-2026-132165 ]
Telegram RCE still going: Earlier this month, security researchers found a way to launch RCE attacks via a new Telegram feature that allows the app to run interactive HTML&JS widgets. This was exploited in the wild and later patched. Researchers have now found a way to bypass the patches and still exploit the feature for RCE attacks. [ NebuSec ]
Pre-auth RCE in Sophos UTM: Security firm Sophos has released a security update for the Sophos Unified Threat Management platform to fix a pre-authentication remote code execution attack. According to security firm Intruder, the vulnerability allows attackers to gain root on the device just by sending an email to an organization running the appliance. Sophos released updates even after it retired the product in June. [ Intruder // Sophos CVE-2026-118633 ]
Nextcloud RCE: Security researchers say a single maliciously crafted HEIC image shared via a public Nextcloud link can trigger unauthenticated Remote Code Execution (RCE) on servers running the official Nextcloud Docker container. [ Pluto Security ]
Unpatched Synology bug: Synology warned customers a bug that can read files from their DSM network attached storage devices. A patch is not available yet. [ Synology CVE-2026-120382 ]
KEV update: CISA has updated its KEV database ( twice ) with five vulnerabilities that are currently exploited in the wild. This includes:
CVE-2024-10441 – Synology DiskStation Manager (DSM) Improper Encoding or Escaping of Output Vulnerability
CVE-2026-40217 – BerriAI LiteLLM Unprotected Alternate Channel Vulnerability
CVE-2026-90436 – Apache Axis Unsafe Reflection Vulnerability
CVE-2026-90817 – Vanderbilt University Research Electronic Data Capture (REDCap) Code Injection Vulnerability
CVE-2026-115731 – Perforce Puppet Enterprise Deserialization of Untrusted Data Vulnerability
Phrack 73: … is out !
Acquisition news: Infosec giant Palo Alto Networks has acquired Applied Computer, a leading platform for post-training and inference of open-weight AI models. [ PAN ]
Threat/trend reports: Aikido Security , Beazley Security , Central Bank of Russia , Flashpoint , Gartner , Iru , NordPass , OTCC , Secure Code Warrior , Trellix , and VanishID have recently published reports and summaries covering various emerging threats and industry trends.
Risky Business podcasts
In this edition of Seriously Risky Business , Tom Uren and Amberleigh Jack talk the Trump administration’s AI task force asking for US federal regulatory agencies to go easy on organisations that fall victim to AI-powered cyber attacks. This makes sense, at least in the short term. But we’d like to see more expansive forward-looking policy that encourages experimentation with AI for defensive purposes.
The full story
This article is one source in a clustered incident — the cluster page carries the summary, timeline and every other outlet covering it.
