November 19, 2026 USD 45.20EUR 49.57PLN 11.24
Новини Сміла — коротко й по суті
LIVE
Powerful Explosions near Crimean Bridge: What It Means for the Region Frontline situation: over 200 combat engagements reported in a day US and Venezuela Announce Enhanced Defense Partnership Germany prepares financial 'safety net' for Ukraine amid potential EU political shifts South Korean defense startup denies reports of halting operations in Ukraine Ukrzaliznytsia to update systems: ticket sales to be temporarily suspended Oatmeal in your daily diet: health benefits and expert advice Ukrzaliznytsia announces temporary suspension of ticket sales Vitaliy Kim on veteran adaptation risks: a debate on social integration Rare emergence of isolated Awa tribe from Amazon jungle: causes and risks Powerful Explosions near Crimean Bridge: What It Means for the Region Frontline situation: over 200 combat engagements reported in a day US and Venezuela Announce Enhanced Defense Partnership Germany prepares financial 'safety net' for Ukraine amid potential EU political shifts South Korean defense startup denies reports of halting operations in Ukraine Ukrzaliznytsia to update systems: ticket sales to be temporarily suspended Oatmeal in your daily diet: health benefits and expert advice Ukrzaliznytsia announces temporary suspension of ticket sales Vitaliy Kim on veteran adaptation risks: a debate on social integration Rare emergence of isolated Awa tribe from Amazon jungle: causes and risks

GPT-6 AI Uncovers Critical Java Vulnerability: NASA and Global Corporations at Risk

💡 Tech 2 min read 👁 3

OpenAI's GPT-6 Cyber model has identified a dangerous flaw in the Java programming language, affecting critical systems worldwide.

GPT-6 AI Uncovers Critical Java Vulnerability: NASA and Global Corporations at Risk💡Tech

Photo: RBC-Ukraine

Key points

  • GPT-6 Cyber found a critical flaw in Java's base64 decoder, named Java4Shell.
  • The vulnerability affects major entities like NASA, CERN, and Fortune 500 firms.
  • AI tools have reduced the time needed for successful system exploitation to under an hour.

The cybersecurity landscape has shifted dramatically following the discovery of a critical vulnerability in Java, identified by OpenAI's specialized GPT-6 Cyber model. Java serves as the backbone for government infrastructure and approximately 90% of Fortune 500 companies. The flaw, dubbed Java4Shell, potentially allows hackers to gain unauthorized access to high-security systems globally.

Technical Breakdown and Impact

Security researcher Quang Luong discovered the issue within the base64 decoder—a standard algorithm for converting binary data into text. It was revealed that under specific conditions, the system fails to correctly process a sequence of 64 'A' characters followed by an equals sign. Utilizing AI tools, the Calif startup team demonstrated that this glitch could be exploited to corrupt computer memory and seize system control.

The threat was validated through a successful test attack on Keycloak, a widely used identity and access management system. Keycloak is utilized by thousands of organizations, including NASA, CERN, and Hewlett Packard. Independent verification showed that an AI-assisted attack could be executed in less than an hour, highlighting a significant decrease in the time required for cybercriminals to strike.

The Rise of 'Centaur Hackers'

Tai Duong, CEO of Calif, noted that generative models are fundamentally altering the speed of both cybersecurity defense and offense. Tasks that previously took large security teams months to complete are now being handled by small groups in a matter of hours. Duong suggests that the most effective operators in this new era are 'centaur hackers'—professionals who work in tandem with AI to amplify their capabilities.

While Oracle has already released a patch to fix the vulnerability, the U.S. Cybersecurity and Infrastructure Security Agency (CISA) has urged organizations to update their systems immediately. Experts warn that the risk will persist for 5 to 10 years for legacy systems that do not receive regular maintenance. This discovery underscores the urgent need for proactive security measures in an increasingly AI-driven threat environment.

Source RBC-Ukraine — rbc.ua/ukr/news/gpt-6-cyber-znayshla-nebezpechnu-vrazlivist-1795001… ↗

Contact the newsroom

Got a tip, a proposal or a partnership question? Write to us and we will reply shortly.