AI has discovered an issue with Java that leaves many companies and other institutions vulnerable.
Using OpenAI’s still-not-public GPT-6 Cyber model, cybersecurity startup Calif found a string of “A” characters followed by an equals sign. While this normally leads to a crash, bad actors can leverage this to gain full unauthorized access to a device.
This is particular dangerous since Java is a ubiquitous programming language used by 90 per cent of Fortune 500 companies, including Amazon and Apple, as well as U.S. banks like J.P. Morgan Chase and the Bank of America and even hospitals.
Calif, for its part, says it’s not easy to use the bug for malicious purposes. Still, it’s an understandably big concern, and the agency says it’s been coordinating with associates in U.S., Canada, Australia, New Zealand and Britain accordingly. Oracle, which owns Java, hasn’t issues a public statement on the matter. Instead, it quietly rolled out a software patch on Tuesday.
This is the latest in a lineup of high-profile tech failures. Last week, Microsoft services like Outlook and Teams weren’t working for many people, which the company attributed to a “latent defect” from a server fix. The same week, U.S. airlines were grounded for hours due to air traffic control system errors.
Naturally, this has led experts to scrutinize the increasing role of AI in both tech businesses and those behind cyberattacks.
News
News
News
News