In a startling revelation, Google declared that advancements in artificial intelligence have enabled programs to autonomously identify and exploit security vulnerabilities in software, often without human intervention. The company’s Threat Intelligence Group asserted that “the era of autonomous attacks has arrived.”
Researchers identified a tool known as exploit_agent operating from an unsecured server. This tool employs sophisticated AI algorithms to monitor software updates closely.
When a company issues a security patch, the tool compares the previous version with the latest, determines the flaw that was fixed, generates the necessary exploit code, and attempts to breach the system autonomously.
The same server contained stolen data from victims, along with detailed procedures for post-breach actions, such as network exploration and implanting backdoors. The activity was primarily aimed at French sites in the media, education, and entertainment sectors, with some notes found in Russian.
Google characterized this development as significant, emphasizing that attacks could unfold at an unprecedented pace. When a security fix is issued, unprotected systems can be targeted almost instantaneously, without requiring an expert to oversee the operation.
Additionally, the report highlighted that other AI-driven tools are now being utilized to scan for vulnerabilities and quickly execute complex attacks. Google, however, did not link these findings to the recent technological disruptions that have captured media attention.
POLL: Will America Regret Giving The Democrats Power?
Hackers are now using LLMs to continuously scan software updates, spot newly fixed bugs, and build attacks before teams can apply the fix — all without human involvement. That’s according to the latest report from the Google Threat Intelligence Group.
We're helping defenders…
— News from Google (@NewsFromGoogle) November 16, 2026
Among these disruptions were a prolonged outage at Microsoft affecting email and workplace functions, an incident at the Federal Aviation Administration that halted flights, delays at United Airlines, and JPMorgan Chase experiencing transaction slowdowns. Both United and JPMorgan determined that their issues were not the result of hacking.
Around this time, a known website-testing service, Optimizely, was compromised. Attackers embedded malicious code into projects used by at least 178 organizations, impacting visitors on sites like Zoom, CNN, the BBC, and HSBC, potentially exposing them to threats aimed at stealing personal information or conducting further attacks on devices.
In response to public apprehension, Vice President JD Vance addressed concerns in a Fox News segment. While questioned about preparedness for potential crises due to recent outages, he reassured that there’s no immediate danger but emphasized the importance of always being prepared.
