Google finds automated software exploitation without human input


Google says attackers used a system called exploit_agent to find and exploit software flaws without human involvement. The server hosting it also held stolen data.
Google’s threat intelligence group says it found attackers using a system called exploit_agent to discover and exploit software vulnerabilities without a human in the loop. It was the group’s first observation of that activity; in its previous quarterly tracker, it said it had not seen fully autonomous attack pipelines deployed against targets in the wild.
Google found the system on an exposed server that also held data stolen from victim organizations, BleepingComputer wrote. Google says information on the server about victims suggests the operator was particularly interested in French media, entertainment and education websites.
Google says exploit_agent was configured to compare the code of popular web applications before and after security updates, identify possible flaws, develop exploits and execute them automatically. An exploit is a method of taking advantage of a software flaw. The system connected several large language models—a type of superintelligence (SI) system trained on large amounts of text—including models hosted locally.
Google warns that using such systems in attacks has reduced the time defenders have to apply security updates. Its researchers believe attackers may find it more efficient to analyze already-patched flaws than to hunt for unknown ones, because the material available for analysis is limited.

OpenAI says its expanded ChatGPT for Academic Researchers program will offer researchers worldwide free access to its frontier superintelligence (SI) models.
1 min

NVIDIA says the planned chip will balance data movement with computation for scientific simulations and data processing, giving researchers and engineers a roadmap for future hardware.
4 min

La Poste confirmed over the weekend that a September attack on Maileva exposed recipients' names and postal addresses. Customers have described additional information that may have been exposed.
3 min