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“.

Company didn’t notice its chatbot was being abused for (at least) 4 months.

OpenAI on Tuesday announced the launch of ChatGPT for government agencies in the U.S. ...It allows government agencies, as customers, to feed “non-public, sensitive information” into OpenAI’s models while operating within their own secure hosting environments, OpenAI CPO Kevin Weil told reporters during a briefing Monday.

The defendants used stolen API keys to gain access to devices and accounts with Microsoft’s Azure OpenAI service, which they then used to generate “thousands” of images that violated content restrictions.

Since the launch of ChatGPT, OpenAI has sparked significant interest among both businesses and cybercriminals. While companies are increasingly concerned about whether their existing cybersecurity measures can adequately defend against threats curated with generative AI tools, attackers are finding new ways to exploit them. From crafting convincing phishing campaigns to deploying advanced credential harvesting and malware delivery methods, cybercriminals are using AI to target end users and capitalize on potential vulnerabilities.

Barracuda threat researchers recently uncovered a large-scale OpenAI impersonation campaign targeting businesses worldwide. Attackers targeted their victims with a well-known tactic — they impersonated OpenAI with an urgent message requesting updated payment information to process a monthly subscription.

We banned accounts linked to an Iranian influence operation using ChatGPT to generate content focused on multiple topics, including the U.S. presidential campaign. We have seen no indication that this content reached a meaningful audience.

Early last year, a hacker gained access to the internal messaging systems of OpenAI, the maker of ChatGPT, and stole details about the design of the company’s A.I. technologies.

The hacker lifted details from discussions in an online forum where employees talked about OpenAI’s latest technologies, according to two people familiar with the incident, but did not get into the systems where the company houses and builds its artificial intelligence.

OpenAI updated its ChatGPT macOS app on Friday after users discovered it stored conversations insecurely in plain text.

A NewsGuard audit found that chatbots spewed misinformation from American fugitive John Mark Dougan.
#AI #Axios #ChatGPT #Google #Illustrations #License #Microsoft #Misinformation #OpenAI #Visuals #genAI #generative #or

OpenAI has appointed Paul M. Nakasone, a retired general of the US Army and a former head of the National Security Agency, to its board of directors.

Covert propagandists have already begun using generative artificial intelligence to boost their influence operations.

When OpenAI CEO Sam Altman announced GPTs, custom chatbots powered by OpenAI's generative AI models, onstage at the company's first-ever developer

Security researchers created an AI worm in a test environment that can automatically spread between generative AI agents—potentially stealing data and sending spam emails along the way.

We terminated accounts associated with state-affiliated threat actors. Our findings show our models offer only limited, incremental capabilities for malicious cybersecurity tasks.

Millions of articles from The New York Times were used to train chatbots that now compete with it, the lawsuit said.

The European Union agreed on terms of the AI Act, a major new set of rules that will govern the building and use of AI and have major implications for Google, OpenAI, and others racing to develop AI systems.

The company later restored access to the chatbot, which is owned by OpenAI.

OpenAI managed to appease Italian data authorities and lift the country’s effective ban on ChatGPT last week, but its fight against European regulators is far from over. 

ChatGPT for a long time on March 20th posted a giant orange warning on top of their interface that they’re unable to load chat history.

Attackers are distributing malware disguised as a ChatGPT desktop client for Windows offering “precreated accounts”