Man with bipolar disorder sued OpenAI after surviving ChatGPT-linked suicide attempt.

Threat actors have been observed using seemingly legitimate artificial intelligence (AI) tools and software to sneakily slip malware for future attacks on organizations worldwide. According to Trend Micro, the campaign is using productivity or AI-enhanced tools to deliver malware targeting various regions, including Europe, the Americas, and the Asia, Middle East, and Africa (AMEA) region. Manufacturing, government, healthcare, technology, and retail are some of the top sectors affected by the attacks, with India, the U.S., France, Italy, Brazil, Germany, the U.K., Norway, Spain, and Canada emerging as the regions with the most infections, indicating a global spread.

EvilAI uses signed AI apps to spread malware globally, stealing data and evading detection.

 A new wave of AI tools developed with no ethical restrictions is allowing hackers to detect and exploit software vulnerabilities faster than ever before. As these "evil AI" platforms advance quickly, cybersecurity experts fear that traditional defences will fail to keep up. Earlier this week at the annual RSA Conference in San Francisco, a crowded room at Moscone Centre assembled for what was touted as a technical investigation of artificial intelligence's involvement in contemporary hacking. The event, conducted by Sherri Davidoff and Matt Durrin of LMG Security, promised more than just theory; it would include a rare, live demonstration of so-called "evil AI" in operation, a topic that has quickly progressed from cyberpunk fiction to real-world concerns. The CEO and founder of LMG Security, Davidoff, opened with a sobering reminder of the constant threat posed by software flaws. According to PCWorld senior editor Alaina Yee, Durrin, the company's Director of Training and Research, swiftly changed the tone. He popularised the idea of "evil AI"—artificial intelligence tools created without moral boundaries that can spot and take advantage of software vulnerabilities before defences can respond. "What if hackers utilise their malevolent AI tools, which lack safeguards, to detect vulnerabilities before we have the opportunity to address them?" Durrin asked the audience, previewing the unsettling demonstrations to come.  The team's attempts to acquire one of these rogue AIs, such as GhostGPT and DevilGPT, frequently resulted in irritation or discomfort. Finally, their persistence paid off when they discovered WormGPT, a tool mentioned in a Brian Krebs piece, for $50 via Telegram channels. As Durrin explained, WormGPT is effectively ChatGPT without the ethical constraints. It will respond to every question, regardless of how harmful or illegal the request. However, the presenters emphasised that the main concern is not the tool's presence, but rather its capabilities.  The LMG Security team began by running an older version of WormGPT through DotProject, an open-source project management platform. The AI accurately discovered a SQL vulnerability and offered a simple exploit, but it failed to construct a viable assault, most likely due to its inability to parse the entire codebase. A revised version of WormGPT was then entrusted with investigating the famed Log4j issue. This time, the AI not only discovered the issue, but also gave enough information that, as Davidoff noted, "an intermediate hacker" could utilise it to craft an exploit.  The true surprise came with the most recent iteration: WormGPT provided step-by-step instructions, complete with code specific to the test server, and those instructions worked beautifully. To test the restrictions further, the team created a susceptible Magento e-commerce platform. WormGPT discovered a complicated two-part exploit that was undetected by popular security products such as SonarQube and ChatGPT itself. During the live demonstration, the rogue AI provided a full hacking guide unprompted and with alarming speed. As the discussion came to a close, Davidoff remarked on the rapid progress of malicious AI tools.

A sophisticated malware campaign, dubbed “EvilAI” by cybersecurity firm Trend Micro, is leveraging AI-enhanced software to infiltrate organizations across the globe. By posing as legitimate productivity tools, this malware evades traditional security measures and exploits user trust. The post EvilAI Malware Exploits AI Targets Organizations Worldwide appeared first on Purism.

A new wave of malware disguised as everyday productivity tools has been quietly spreading across the internet, stealing user credentials and giving attackers remote control of infected systems. Researchers have tracked hundreds of campaigns tied to a threat known as TamperedChef, also called EvilAI, which wraps dangerous code inside apps that look and feel completely legitimate. Since early 2023, attackers have packaged malware inside tools like PDF editors, calendar apps, ZIP extractors, and GIF image makers. These apps work as advertised, which is exactly why victims rarely suspect anything at all. They sit silently on a device for weeks or even months before triggering malicious activity, making them difficult to catch with standard security tools. Analysts at Unit42 identified and tracked three distinct clusters of this activity, labeled CL-CRI-1089, CL-UNK-1090, and CL-UNK-1110. According to Unit42 report  shared with Cyber Security News (CSN), researchers found over 4,000 unique samples and more than 100 unique variants across these campaigns, with infections appearing in more than 50% of monitored enterprise environments globally. What makes TamperedChef so dangerous is how convincingly it mimics real software. Download pages are professionally built with legal terms, contact pages, and one-click download buttons on legitimate-looking domains. TamperedChef Malware Uses Signed Productivity Apps The apps deliver on their promises, leaving victims with little reason to question what they just installed. The scale of this operation points to a well-funded, highly organized effort. Researchers estimate the operators behind just one cluster spent over $10,000 on code-signing certificates alone, which are digital stamps that make software appear trustworthy. This level of investment signals a long-term, profit-driven campaign far beyond what typical adware operations would attempt. One of TamperedChef’s defining tactics is using legitimate code-signing certificates to make its payloads appear safe. These certificates are issued to verified companies, and most security tools treat signed software as trustworthy. Threat actors exploited this by building networks of shell companies across Ukraine, Malaysia, Israel, the UK, and the US to obtain valid certificates. Researchers traced the CL-CRI-1089 cluster to 34 unique code-signing entities, connected through shared certificate usage, overlapping code, and corporate structure analysis. The Calendaromatic campaign used a self-extracting archive containing a functional calendar app bundled with a hidden remote access Trojan. Once active, that RAT contacted a command-and-control server and pulled down a second-stage payload to further compromise the victim. The CL-UNK-1090 cluster took a more integrated approach, with the same group owning both the advertising agencies and the malware-signing companies. Examples of download pages for TamperedChef-style fake productivity applications (Source – Unit42) Over 20,000 unique ads were traced to this cluster through ad transparency platforms, spanning campaigns like CrystalPDF, OneZip, and Easy2Convert. Operators used generative AI to build distribution websites at scale, producing pages that looked similar but had structurally different underlying code. Stealers, RATs, and What Happens After Infection Once a TamperedChef app activates, it delivers one of two payload categories depending on the campaign. The first is adware and browser hijackers, which redirect searches and take control of browsing behavior. Simplified signature flow of reuse between samples (Source – Unit42) The second, and more serious, is the deployment of information stealers and remote access Trojans that target saved credentials and allow attackers to run commands remotely. Second-stage payloads typically arrive weeks after installation through an upstream API connection, long after any initial suspicion fades. In some campaigns, such as AppSuite, researchers also found proxy-style malware routing traffic through victim machines. The CL-CRI-1089 cluster showed the most aggressive credential theft, while CL-UNK-1090 favored stealthier in-memory payloads leaving fewer traces on disk. To defend against this threat, security teams should ensure endpoint detection tools are fully updated across all devices and consider enterprise browsers that block malicious downloads before they reach users. Training employees to recognize unfamiliar software risks is equally critical, even when download sites look entirely professional. If an infection is discovered, teams should quarantine related files, remove persistence mechanisms like scheduled tasks, reset credentials for affected accounts, and review access logs to confirm whether stolen credentials have already been misused. Indicators of Compromise (IoCs):- Type Indicator Description SHA256 Hash 248de1470771904462c91f146074e49b3d7416844ec143ade53f4ac0487fdb4 RapiDoc binary containing PDB path, linked to CANDY TECH LTD (CL-UNK-1090) SHA256 Hash 42231bfa7c7bd4a8ff12568074f83de8e4ec95c226230cccc6616a1a4416de268 RapiDoc binary containing PDB path, linked to CANDY TECH LTD (CL-UNK-1090) PDB Path D:!Work\Clients\<user>\Projects\RapiDoc\SrcForTests\RapiDoc\x64\Release\RapiDoc\RapiDoc.pdb Program database path found in RapiDoc binaries, likely left by mistake during build Domain onezipapp[.]com Distribution site for OneZip malware, signed by TAU CENTAURI LTD (CL-UNK-1090) Domain crystalpdf[.]com Distribution site for CrystalPDF, used by CL-UNK-1090 cluster Domain Pattern pixel.toolname[.]com C2 domain pattern used by PixelCheck variant (PDFPrime/ManualzPDF campaigns, CL-CRI-1089) Code Signer CROWN SKY LLC Code-signing entity used in Calendaromatic campaign (CL-CRI-1089) Code Signer MARKET FUSION INNOVATIONS LLC Code-signing entity linked to Calendaromatic campaign (CL-CRI-1089) Code Signer CANDY TECH LTD Core signing and advertising entity for CL-UNK-1090 cluster Code Signer TAU CENTAURI LTD Signing entity linked to OneZip campaign (CL-UNK-1090) Code Signer B.L.A ASPIRE LTD Signing entity for JustConvertFiles binaries (CL-UNK-1090) Code Signer PASTEL CONCEPTION LTD Signing entity for JustConvertFiles; linked to PDFPilot, SwiftNav, ShinyPDF, FileEase Code Signer BUZZ BOOST ADVERTISERS LLC Certificate entity linked to PixelCheck variant (CL-CRI-1089) Code Signer ADSMARKETO LLC Certificate entity linked to PixelCheck variant (CL-CRI-1089) Code Signer ADVANTAGE WEB MARKETING LLC Certificate entity linked to PixelCheck variant (CL-CRI-1089) Code Signer Europae-Solutio Ltd Certificate entity linked to PixelCheck variant (CL-CRI-1089) Code Signer SP Development and Solution Limited Certificate entity linked to PixelCheck variant (CL-CRI-1089) Code Signer LLC MATCH-TWO-USERS Certificate entity linked to PixelCheck variant (CL-CRI-1089) Code Signer Monetize forward LLC Certificate entity linked to PixelCheck variant (CL-CRI-1089) Malware Sample calendaromatic-win_x64.exe First-stage binary from Calendaromatic campaign (CL-CRI-1089) Malware Sample resources.neu Obfuscated NeutralinoJS resource file containing C2 logic, Calendaromatic campaign File Name RapiDoc.pdb Debug symbol file found in RapiDoc binaries (CL-UNK-1090) Campaign Name AppSuite PDF Malicious PDF editor spreading TamperedChef malware; observed deploying proxy-style payloads Campaign Name Calendaromatic Calendar app trojan; earliest tracked CL-CRI-1089 activity (late 2023) Campaign Name CrystalPDF Malicious PDF tool distributed by CL-UNK-1090; hosted at crystalpdf[.]com Campaign Name JustAskJacky App distributed by CL-UNK-1110 cluster Campaign Name OneZip Malicious ZIP tool signed by TAU CENTAURI LTD; distributed via onezipapp[.]com Campaign Name PDFPrime / ManualzPDF Early CL-CRI-1089 campaigns sharing code and C2 patterns (PixelCheck variant) Campaign Name ZipMakerPro TamperedChef-style app linked to CANDY TECH LTD (CL-UNK-1090) Campaign Name GifsMakerPro TamperedChef-style app linked to CANDY TECH LTD (CL-UNK-1090) Campaign Name ScreensRecorder TamperedChef-style app linked to CANDY TECH LTD (CL-UNK-1090) Campaign Name RapiDoc App with CANDY TECH LTD copyright; contained leaked PDB path (CL-UNK-1090) Campaign Name JustConvertFiles Malicious file conversion tool distributed by CANDY TECH LTD (CL-UNK-1090) Campaign Name PDFPilot Campaign linked to B.L.A ASPIRE LTD and PASTEL CONCEPTION LTD (CL-UNK-1090) Campaign Name SwiftNav Campaign linked to B.L.A ASPIRE LTD and PASTEL CONCEPTION LTD (CL-UNK-1090) Campaign Name ShinyPDF Campaign linked to B.L.A ASPIRE LTD and PASTEL CONCEPTION LTD (CL-UNK-1090) Campaign Name FileEase Campaign linked to B.L.A ASPIRE LTD and PASTEL CONCEPTION LTD (CL-UNK-109 Note:   IP addresses and domains are intentionally defanged (e.g.,  [.] ) to prevent accidental resolution or hyperlinking. Re-fang only within controlled threat intelligence platforms such as MISP, VirusTotal, or your SIEM . Follow us on  Google News ,  LinkedIn , and  X  to Get More Instant Updates ,  Set CSN as a Preferred Source in  Google . The post TamperedChef Malware Uses Signed Productivity Apps to Deliver Stealers and RATs appeared first on Cyber Security News .