Skip to content
Bramble

Technology

Startup uses OpenAI's GPT-6 Cyber to find critical bug in Java

Cybersecurity startup Calif said on 17 November it used OpenAI's GPT-6 Cyber model to find a critical flaw in Java. Oracle released a patch the same day.

2 outlets · 0L · 1C · 1R First reported Account updated
Image: The Straits Times
Image: Anadolu Agency

1 / 2

The story, neutrally told

Mixed · 2Cybersecurity startup Calif said on 17 November that it had used OpenAI's recently released GPT-6 Cyber model to uncover a critical software bug in Java. Mixed · 2Java is used by government agencies and by about 90% of Fortune 500 companies, including in banking, health care, aviation and technology. Mixed · 2Bruce Dang, a senior researcher at Calif, said: "Java is the beating heart of the modern economy," and added that "all major companies" would practically speaking be affected by the bug.

Right · 1According to The New York Times, the flaw could let a hacker gain remote access to sensitive internal systems, and it may be Calif's most serious find yet. Right · 1Dang and colleagues have worked in recent days to coordinate with the Five Eyes governments (the United States, Canada, Australia, New Zealand and Britain) and with major private vendors including Google and Microsoft. Right · 1Oracle, which owns the Java trademark and sells commercial Java licences, released a software patch on 17 November but did not respond to a request for comment.

Right · 1The National Security Agency's classified network went down on 17 November for what it called "routine software maintenance"; the NSA did not comment, and it was not clear whether this was linked to the Java bug. Mixed · 2The discovery is presented as the latest sign that AI is finding flaws in code so fast that security professionals struggle to keep up. Calif has found a string of serious vulnerabilities in recent months using leading AI models. Right · 1The New York Times noted that recent outages of Microsoft products, a nationwide grounding of flights and banking interruptions have fanned fears that AI is to blame, though the explanations so far appear relatively benign.

Every sentence links to the reporting it rests on. The pill in front of each says where its sources sit: Left, Centre or Right when one side supplies at least half of them, Mixed when they are evenly split. The number is how many outlets it cites.

Left0 outlets

No left outlet in our sources has covered this story yet.

Centre1 outlet

Framing
Short report relaying The New York Times: an AI model found a critical Java flaw, showing AI's growing role in vulnerability research.
Emphasis
Java's reach and Dang's warning that all major companies will be affected; concern among experts about AI-powered vulnerability research.
Leaves out or plays down
Omits the Oracle patch, the Five Eyes and vendor coordination, the NSA outage and the recent unexplained outages.
Charged language
“critical security vulnerability”
For example
“The discovery underscores growing concerns among cybersecurity experts and technology companies about the implications of AI-powered vulnerability research.” — Anadolu Agency

Right1 outlet

Framing
Full New York Times article run by The Straits Times: the find as evidence of AI transforming vulnerability discovery faster than defenders can keep up.
Emphasis
Patch, disclosure coordination with Five Eyes and vendors, the NSA network outage and recent outages that fanned AI fears, with the caveat that explanations appear benign.
Leaves out or plays down
No Oracle or NSA comment; the technical details of the flaw are not given.
Charged language
“revolutionising”“startled the cybersecurity world”“wreak havoc”
For example
“It was the latest evidence of a reality facing companies and the government, as AI is revolutionising the field of discovery in computer vulnerabilities” — The Straits Times

Ask about this story

Loading…