US, South Korea issue urgent advisories after OpenAI Cyber model flags critical security flaws in Java | World News
The United States and South Korea have issued urgent cybersecurity advisories after a cybersecurity firm unveiled a critical software bug in Java, a programming language widely used in key information systems, including government agencies, financial institutions and the military.
Java is widely used by banks, hospitals, the aviation industry, government agencies, Fortune 500 companies, the corporate sector and even the military. (Unsplash/Representative Image)According to a report by The New York Times, the start-up Calif exposed a major flaw in Java using OpenAI’s GPT-6 Cyber model. Bruce Dang, a senior researcher at Calif, described Java as the "beating heart of the modern company", adding that vulnerabilities in the programme will affect all major companies.
Cybersecurity concerns around the world have mounted this year with the rapid pace of AI development and tech outages in recent months. Calif and other groups have been using AI to help alert network defenders to vulnerabilities that can be exploited by hackers to carry out a major security breach.
However, the Java discovery is Calif's most serious threat, as the programming language is widely used by banks, hospitals, the aviation industry, government agencies, Fortune 500 companies, the corporate sector and even the military. It also helps power cloud computing architecture, identity management tools and email services.
According to the report, the security flaw in Java could allow a hacker to gain remote access to sensitive internal systems. This discovery has assumed significance after widespread tech disruptions in recent days, including the outages of Microsoft products and the nationwide ground stop of all departing flights in the US due to a "communications failure".
The bug, called Java4Shell, is reportedly caused while trying to execute 64 'A' characters followed by one equal sign. It caused Java to erroneously execute code by tampering with a section of computer memory, which could be used to corrupt data and grant hackers full access to any system.
Calif said they could use this bug to break into Keycloak, a widely used authentication and login system that is associated with thousands of big tech organisations, including Hewlett Packard, UnitedHealthcare and NASA, as per the report.
The Java vulnerability is the second major bug discovered by Calif using OpenAI models. Last month, it exposed security flaws in PHP, a programming language estimated to be used by 70% of websites and apps on the internet, triggering a frantic effort to deploy a patch before government agencies took websites offline to prevent hackers from exploiting the bug.
The US Cybersecurity and Infrastructure Security Agency issued an alert, pressing organisations to “immediately patch” the vulnerability. "Review your software inventory for JDKs and Java runtimes, including those bundled with third-party applications or included in container images and appliances," it said.
In South Korea, which is considered to be particularly vulnerable as it heavily relies on Java in the public and private sectors, the presidential office held an emergency meeting on Tuesday with related government agencies to discuss preemptive measures to tackle the issue, according to the Yonhap news agency.
South Korea's Ministry of Science and ICT and the Korea Internet & Security Agency said on Wednesday that public institutions and private enterprises must urgently install the security updates. Java represented 53% of programming language use in South Korea compared to a global average of 30%, according to a JetBrains survey.
This came amid global fears of AI resurgence that could enable more of such tech vulnerabilities and hacking attempts. Quang Luong, a security researcher from Calif, said it was only a matter of time before AI can independently build the most advanced cyberattacks, leading to several unpredictable problems.
Aveek Banerjee is a Senior Content Producer at the Hindustan Times, based in Greater Noida West. With more than five years of experience in digital journalism, Aveek has worked extensively in covering national politics and international affairs, including elections, high-profile diplomatic engagements and geopolitical conflicts. His experience covers a wide range of major events, including the 2024 Lok Sabha Elections, multiple state assembly elections, US elections, Operation Sindoor, Air India crash, India-US trade tensions to the ongoing conflicts in Ukraine and Iran. Before joining Hindustan Times, Aveek worked at the editorial teams of CNN-News18, India TV and Asian News International (ANI), during which he developed an understanding of the high-pressure newsroom environment and honed his skills in covering breaking news, tracking crucial developments, and producing accurate news content under tight deadlines while maintaining editorial standards. These stints have also provided exposure to diverse newsroom environments and the evolving standards of the media industry. Aveek holds a Master's degree in Global Studies from Ambedkar University Delhi and a Bachelor's degree in journalism and mass communication from Vivekananda Institute of Professional Studies. Outside the newsroom, he enjoys reading, pursuing photography and exploring new places in his free time. You can reach him at aveek.banerjee@htdigital.in.Read More