Google has issued an urgent security update for its Chrome browser, addressing a critical zero-day vulnerability that is being actively exploited by attackers.
The flaw, tracked as CVE-2025-6554, is a type confusion vulnerability in Chrome’s V8 JavaScript engine, which underpins the browser’s ability to process web content across Windows, macOS, and Linux platforms.
The vulnerability was discovered by Clément Lecigne of Google’s Threat Analysis Group (TAG) on June 25, 2025. According to Google, attackers have already developed and deployed exploits targeting this flaw in the wild, prompting the company to act quickly.
Google on Monday released a fresh Chrome 137 update to address three vulnerabilities, including a high-severity bug exploited in the wild.
Tracked as CVE-2025-5419, the zero-day is described as an out-of-bounds read and write issue in the V8 JavaScript engine.
“Google is aware that an exploit for CVE-2025-5419 exists in the wild,” the internet giant’s advisory reads. No further details on the security defect or the exploit have been provided.
However, the company credited Clement Lecigne and Benoît Sevens of Google Threat Analysis Group (TAG) for reporting the issue.
TAG researchers previously reported multiple vulnerabilities exploited by commercial surveillance software vendors, including such bugs in Chrome. Flaws in Google’s browser are often exploited by spyware vendors and CVE-2025-5419 could be no different.
According to a NIST advisory, the exploited zero-day “allowed a remote attacker to potentially exploit heap corruption via a crafted HTML page”. It should be noted that the exploitation of out-of-bounds defects often leads to arbitrary code execution.
The latest browser update also addresses CVE-2025-5068, a medium-severity use-after-free in Blink that earned the reporting researcher a $1,000 bug bounty. No reward will be handed out for the zero-day.
The latest Chrome iteration is now rolling out as version 137.0.7151.68/.69 for Windows and macOS, and as version 137.0.7151.68 for Linux.
In this post I’ll show you how I found a zeroday vulnerability in the Linux kernel using OpenAI’s o3 model. I found the vulnerability with nothing more complicated than the o3 API – no scaffolding, no agentic frameworks, no tool use.
Recently I’ve been auditing ksmbd for vulnerabilities. ksmbd is “a linux kernel server which implements SMB3 protocol in kernel space for sharing files over network.“. I started this project specifically to take a break from LLM-related tool development but after the release of o3 I couldn’t resist using the bugs I had found in ksmbd as a quick benchmark of o3’s capabilities. In a future post I’ll discuss o3’s performance across all of those bugs, but here we’ll focus on how o3 found a zeroday vulnerability during my benchmarking. The vulnerability it found is CVE-2025-37899 (fix here), a use-after-free in the handler for the SMB ‘logoff’ command. Understanding the vulnerability requires reasoning about concurrent connections to the server, and how they may share various objects in specific circumstances. o3 was able to comprehend this and spot a location where a particular object that is not referenced counted is freed while still being accessible by another thread. As far as I’m aware, this is the first public discussion of a vulnerability of that nature being found by a LLM.
Before I get into the technical details, the main takeaway from this post is this: with o3 LLMs have made a leap forward in their ability to reason about code, and if you work in vulnerability research you should start paying close attention. If you’re an expert-level vulnerability researcher or exploit developer the machines aren’t about to replace you. In fact, it is quite the opposite: they are now at a stage where they can make you significantly more efficient and effective. If you have a problem that can be represented in fewer than 10k lines of code there is a reasonable chance o3 can either solve it, or help you solve it.
Benchmarking o3 using CVE-2025-37778
Lets first discuss CVE-2025-37778, a vulnerability that I found manually and which I was using as a benchmark for o3’s capabilities when it found the zeroday, CVE-2025-37899.
CVE-2025-37778 is a use-after-free vulnerability. The issue occurs during the Kerberos authentication path when handling a “session setup” request from a remote client. To save us referring to CVE numbers, I will refer to this vulnerability as the “kerberos authentication vulnerability“.
A vulnerability has been identified and remediated in all supported versions of the Commvault software. Webservers can be compromised through bad actors creating and executing webshells.
Exploiting this vulnerability requires a bad actor to have authenticated user credentials within the Commvault Software environment. Unauthenticated access is not exploitable. For software customers, this means your environment must be: (i) accessible via the internet, (ii) compromised through an unrelated avenue, and (iii) accessed leveraging legitimate user credential
For the third time in as many months, Apple has released an emergency patch to fix an already exploited zero-day vulnerability impacting a wide range of its products.
The new vulnerability, identified as CVE-2025-24201, exists in Apple's WebKit open source browser engine for rendering Web pages in Safari and other apps across macOS, iOS, and iPadOS. WebKit is a frequent target for attackers because of how deeply integrated it is with Apple's ecosystem.
Did you know there’s widespread exploitation of FortiNet products going on using a zero day, and that there’s no CVE? Now you do.
A Telegram for Android zero-day vulnerability dubbed 'EvilVideo' allowed attackers to send malicious Android APK payloads disguised as video files.
Sources suspect China is behind the targeted exploitation of two zero-day vulnerabilities in Cisco’s security appliances.
In April 2021 I participated in Pwn2Own Vancouvver competition as a single player, and successfully demonstrated a 0-day virtual machine escape exploit with code execution on Parallels hypervisor. Today I am finally releasing the exploit source code together with a technical walkthrough video talk that I gave on Zero Day Engineering livestream in November 2021.
Starting January 10, 2024, multiple parties (Ivanti, Volexity, and Mandiant) disclosed the existence of a zero-day exploit chain affecting Ivanti Connect Secur…
Volexity has uncovered active in-the-wild exploitation of two vulnerabilities allowing unauthenticated remote code execution in Ivanti Connect Secure VPN appliances. An official security advisory and knowledge base article have been released by Ivanti that includes mitigation that should be applied immediately. However, a mitigation does not remedy a past or ongoing compromise. Systems should simultaneously be thoroughly analyzed per details in this post to look for signs of a breach.
Chinese threat actors exploited a new zero-day vulnerability in Barracuda's Email Security Gateway (ESG) appliances.
Data for almost 36 million customers now in the hands of unknown hackers.
Internet scans show 7,000 devices may be vulnerable. The true number could be higher.
TAG’s discovery of a 0-day exploit used to steal email data from international government organizations.
CVE-2023-47246, a zero-day vulnerability in SysAid IT service management software has been exploited by Cl0p ransomware affiliates.
(0Day) Microsoft Exchange ChainedSerializationBinder Deserialization of Untrusted Data Remote Code Execution Vulnerability
A critical severity vulnerability impacting Royal Elementor Addons and Templates up to version 1.3.78 is reported to be actively exploited by two WordPress security teams.
On 10/4/2023, Atlassian published a security advisory on CVE-2023-22515, a privilege escalation vulnerability affecting Confluence Server & Data Center.
The three zero-day flaws addressed by Apple on September 21, 2023, were leveraged as part of an iPhone exploit chain in an attempt to deliver a spyware strain called Predator targeting former Egyptian member of parliament Ahmed Eltantawy between May and September 2023.
"The targeting took place after Eltantawy publicly stated his plans to run for President in the 2024 Egyptian elections," the Citizen Lab said, attributing the attack with high confidence to the Egyptian government owing to it being a known customer of the commercial spying tool.
