Attackers Abuse ChatGPT Custom GPTs to Deliver RAT Via ClickFix Lures

, , ,

Threat actors are abusing ChatGPT's Custom GPT feature to trick users into visiting malicious sites that deploy ClickFix lures, leading to the download of a remote access trojan (RAT). The attack starts with sponsored search results directing victims to attacker-created Custom GPTs, which provide links to fake Cloudflare CAPTCHA pages that execute PowerShell commands to install malware using DLL sideloading and evasion techniques. Over 40 users have been infected so far in this campaign, highlighting continued exploitation of trusted AI platforms for social engineering and malware delivery.

https://thehackernews.com/2026/09/attackers-abuse-chatgpt-custom-gpts-to.html

TeamViewer Urges Users to Patch Severe Flaws “as Soon as Possible”

TeamViewer has released security updates to address multiple high-severity vulnerabilities, including a remote session access control bypass that could allow unauthorized remote code execution across Windows, Linux, and macOS platforms. The company strongly urges users to update to version 15.82 immediately to mitigate risks, although no active exploitation or public exploit code has been detected so far. These flaws pose risks of code execution and privilege escalation, highlighting the importance of prompt patching to maintain system security.

https://www.bleepingcomputer.com/news/security/teamviewer-urges-users-to-patch-severe-flaws-as-soon-as-possible/

Apple Patches CoreGraphics Zero-Day Flaw Exploited in Attacks

,

Apple has released security updates to fix a zero-day vulnerability (CVE-2026-86950) in its CoreGraphics framework, exploited in sophisticated targeted attacks on iOS devices. The flaw, an out-of-bounds write weakness, could allow arbitrary code execution, and Apple addressed it with improved bounds checking in iOS 26.7.1, iPadOS 26.7.1, and macOS updates for impacted devices. Users are urged to install these patches promptly to prevent potential exploitation.

https://www.bleepingcomputer.com/news/security/apple-patches-coregraphics-zero-day-flaw-exploited-in-attacks/

JadePuffer Crims Hijacked Azure Identities and Used Them to Blow up Cloud Resources

, , , ,

Microsoft has identified the cybercriminal group Storm-3168, associated with the JadePuffer agentic ransomware, hijacking Azure service principals to conduct extensive cloud resource reconnaissance, destruction, and credential theft over an 18-hour period. The attackers deleted numerous Azure Storage accounts, Key Vaults, and other resources while targeting backup and recovery mechanisms, suggesting preparation for a ransomware attack, although no ransom demand or confirmed data exfiltration occurred. This incident highlights risks from exposed credentials and the potential for automated, AI-driven ransomware operations within cloud environments.

https://www.theregister.com/security/2026/09/28/jadepuffer-crims-hijacked-azure-identities-and-used-them-to-blow-up-cloud-resources/5299591

AI-Powered Phishing Platform EvilTokens Disrupted by Microsoft

, ,

Microsoft has disrupted EvilTokens, an AI-driven phishing platform responsible for compromising over 12,000 email accounts across 10,000 organizations worldwide through device code phishing. The platform used AI to craft targeted phishing messages, manage compromised inboxes, and optimize attacks, operating as a paid service until Microsoft seized 50 related websites, disabled 150 domains, and helped arrest two suspects in the UK.

https://www.securityweek.com/ai-powered-phishing-platform-eviltokens-disrupted-by-microsoft/

How Device Code Phishing Gives Scammers Access to Your Account

,

Device code phishing exploits the OAuth 2.0 Device Authorization Grant by tricking victims into entering a legitimate short code on a real sign-in page, unknowingly approving a scammer's sign-in request and granting them access to the victim’s account. This attack bypasses typical security measures such as multifactor authentication because the victim themselves authorizes the sign-in, allowing attackers to obtain authentication tokens to access emails, files, or other services depending on granted permissions. Users should be wary of unexpected requests to enter sign-in codes purportedly for meetings or documents and verify legitimate device sign-ins, while monitoring account activity if a scam is suspected.

https://www.malwarebytes.com/blog/how-to/2026/09/how-device-code-phishing-gives-scammers-access-to-your-account

AI Coding Agents’ 0-Click RCE Flaw Could Hand Attackers Keys to the Kingdom

, ,

A zero-click remote code execution vulnerability called Plugin4Shell affects major AI coding agents—including Anthropic’s Claude Code, OpenAI’s Codex, Google’s Gemini CLI, and Microsoft’s Copilot—by exploiting a flaw in how these agents enforce SHA-pinning for marketplace plugins. This supply-chain attack allows attackers to replace trusted plugins with malicious versions without user interaction, potentially granting full access to affected systems. While Anthropic and OpenAI have patched the flaw, Google and Microsoft have not fully addressed it, leaving users vulnerable.

https://www.theregister.com/security/2026/09/17/ai-coding-agents-0-click-rce-flaw-could-hand-attackers-keys-to-the-kingdom/5297335

Claude Opus 5 Helped Researchers Take Over OpenAI Staff Accounts Via Chained Flaws

, , ,

Researchers at Hacktron leveraged Anthropic’s Claude Opus 5 to chain vulnerabilities—starting from a memory-corrupting image flaw in OpenAI’s public forum software to weaknesses in OpenAI’s single sign-on system—enabling them to take over OpenAI staff ChatGPT and Codex accounts and gain access to internal code repositories. The team reported the issues responsibly, leading to a fix within 14 hours and a $6,500 bounty, while highlighting broader risks where shared single sign-on between public and internal systems can escalate compromises. They also demonstrated AI-assisted exploit development, underscoring how advanced models are accelerating offensive security research and attack capabilities.

https://thehackernews.com/2026/09/claude-opus-5-helped-researchers-take.html

Public Exploits Released for Four Linux Kernel Flaws That Enable Local Root

,

A security researcher has publicly released exploits for four recently patched Linux kernel vulnerabilities—DirtyAH6, TUNderflow, PPPoEject, and DiagSpill—that enable local users to escalate privileges to root. The flaws affect various networking components and primarily require unprivileged user namespaces enabled, except DiagSpill, which needs no special privileges if SCTP is present. Users are urged to update to patched kernel versions or disable unprivileged namespaces and affected features to mitigate risk.

https://thehackernews.com/2026/09/public-exploits-released-for-four-linux.html

First Agentic AI Data Breach Reported to Spanish Regulator

,

The Spanish Data Protection Agency (AEPD) reported the first known personal data breach executed through an autonomous AI agent that chained together login, vulnerability scanning, and unauthorized data access. This incident marks a shift in cyber threats, highlighting the need to integrate AI-assisted defense mechanisms, improve incident response times, and strengthen digital identity protections. While the exact cause remains under investigation, the breach demonstrates the growing real-world risks posed by agentic AI in cybersecurity.

https://www.securityweek.com/first-agentic-ai-data-breach-reported-to-spanish-regulator/

Scammers Are Watching Airline Complaints and Posing as Customer Support

, , ,

Check Point researchers uncovered a widespread scam where threat actors monitor public airline complaints on social media, impersonate customer support accounts, and engage victims in private chats to steal personal and payment information under the guise of issuing refunds or compensation. Thousands of fake accounts were identified across platforms like X, Facebook, and Instagram, with hundreds more emerging daily, employing evolving tactics often shifting conversations to WhatsApp. The campaign illustrates a growing external exposure risk for brands and highlights the need for organizations to monitor impersonation, educate customers, and coordinate with platforms to mitigate fraud.

https://blog.checkpoint.com/exposure-management/scammers-are-watching-airline-complaints-and-posing-as-customer-support/

N0va Phishkit Targets US and EU Businesses: a New Challenge for Identity Security

, , ,

The N0va phishkit is conducting sophisticated phishing campaigns targeting organizations across North America and Europe, leveraging trusted business platforms and legitimate authentication flows to steal access tokens and gain single sign-on entry to corporate resources. This identity compromise risks financial loss, sensitive data exposure, operational disruption, and reputational damage, especially in high-risk sectors such as government, healthcare, and technology. Security teams are advised to use advanced threat intelligence tools and behavioral analysis to detect, contextualize, and respond swiftly to N0va-related attacks to minimize impact and contain breaches effectively.

https://thehackernews.com/2026/09/n0va-phishkit-targets-us-and-eu.html

Mythos Has Made 2026 Patching Hell. It Might Make 2027 a Breeze

,

Gartner research vice president Craig Lawson suggests that the surge in vulnerability discoveries in 2026, driven by AI tools like Anthropic’s Mythos, is helping to uncover and address much of the technical debt in large software codebases. This extensive auditing could lead to a decline in the number and severity of vulnerabilities in 2027, as vendors use AI not only to find flaws in existing software but also to improve testing of future releases. Additionally, AI-driven bug hunting and automated remediation may enable more frequent and efficient security assessments, potentially easing the patching burden for cybersecurity teams.

https://www.theregister.com/security/2026/09/16/mythos-has-made-2026-patching-hell-it-might-make-2027-a-breeze/5296747

GitLab’s Critical Flaw Is Already Drawing Internet-Wide Probes

,

GitLab released emergency patches addressing two critical vulnerabilities, including a severe unauthenticated file read flaw (CVE-2026-85706) rated CVSS 10.0, that affects multiple versions of its Community and Enterprise Editions. Security firm WatchTowr Labs has already observed active exploitation attempts, prompting GitLab to urge immediate upgrades for self-managed installations, while CISA added the vulnerabilities to its Known Exploited Vulnerabilities list.

https://cyberscoop.com/gitlab-critical-flaws-path-traversal-scans/

Hackers Used Claude to Hunt for Secrets in 1.8 Million Android Apps

, ,

Anthropic's latest threat intelligence report reveals that hackers used its AI model Claude to automate large-scale cyber operations, including harvesting credentials from 1.8 million Android apps and developing malware and phishing tools. The report highlights that AI is becoming an operational force across malicious activities such as cybercrime, surveillance, propaganda, fraud, and weapons development, enabling even small groups to conduct sophisticated campaigns at scale with less human intervention. These findings underscore a significant shift in the economics and speed of cyber threats, where AI-driven automation multiplies attackers’ capabilities and complicates defense efforts.

https://securityaffairs.com/198905/ai/anthropic-ai-misuse-is-entering-a-new-phase-from-cybercrime-to-surveillance-propaganda-and-weapons.html

Scroll to Top