← Monday, November 16 deck

05Security

Google Threat Intelligence, via Polymarket: The Era of Autonomous Attacks Has Arrived

JUST IN: Google declares “The era of autonomous attacks has arrived,” warning that AI is letting hackers develop attacks faster & at greater scale.

Context

Why it matters

If attackers can study software patches and build exploits with no person in the loop, the time between a fix and a working attack shrinks, and defenders cannot keep working only at human speed.

Polymarket posted that Google declared the era of autonomous attacks has arrived, warning that AI is letting hackers develop attacks faster and at greater scale. An official Google blog post by Google Threat Intelligence Group uses that same headline and summarizes a new AI Threat Tracker report. An official Google Cloud Security post about the report had only about 1,200 views, so it did not meet the 100,000 view floor for a lead. The Polymarket post is the lead, with the official blog and company post in the links.

Deep dive

Google Threat Intelligence Group published a blog post on November 16 titled The era of autonomous attacks has arrived. The post says that, according to the latest AI Threat Tracker report, attackers are now using autonomous large language model frameworks to scan software updates, work out which flaw a patch fixed, and build an attack, all without human involvement. Google wrote that defending against machine speed threats at human speed is not feasible, and it pointed to a product called Google AI Threat Defense that combines Gemini and other models, Wiz, CodeMender, and Mandiant.

Secondary reporting on the report describes a framework GTIG calls exploit_agent, found on an exposed staging server, that compares code before and after a patch, drafts an exploit, and runs it. That reporting also says GTIG had not observed a fully autonomous attack pipeline in the wild in its prior edition two months earlier. Those details are attributed to coverage of the report and should be checked against the full tracker document.

The Polymarket lead is a one line summary and uses an HTML ampersand in the post text, which is kept exactly as the X API returned it. It is not a Google account, which is why the official blog is the source of record for the wording.

What to expect next

The full AI Threat Tracker report is the document to read for case counts, the exploit_agent details, and how far these pipelines go beyond writing exploits. Watch whether other labs confirm similar activity and whether Google AI Threat Defense is offered with clear availability and pricing. Defenders will also be watching how quickly n day exploits appear after patches are released.