Google Warns AI Cyber Threats Cross Key Threshold
Google says criminals increasingly use autonomous AI agents to find software flaws and build exploits, as several major firms reported separate outages this week.
Updated on Nov. 17, 2026 in Cybersecurity

Live Poll
Do you trust airlines and banks to keep their systems running reliably?
Google published a threat-intelligence report on Monday warning that criminals increasingly deploy autonomous AI agents to identify software vulnerabilities and develop exploits, stating that the era of autonomous attacks has arrived. The warning came as United Airlines, JPMorgan Chase, and Microsoft each experienced technology disruptions, though neither Google nor the affected companies has linked the threats to those outages.
Why it matters
Google's report suggests that attacks able to operate without continuous human involvement could shorten the time defenders have to find and patch flaws. The recent outages highlight how exposed critical systems are, even though investigators so far have found no evidence of cyberattacks behind them.
Google's report describes autonomous AI agents that identify software vulnerabilities, develop exploits, and attack systems without continuous human involvement. The FAA failure was traced to malfunctioning routers following network upgrades, but whether AI played any role in the outages has not been determined.
The players
Google is a technology company whose research and threat-intelligence teams track cyber threats across global software and cloud systems.
United Airlines
United Airlines is a major U.S. carrier that operates one of the largest airline networks in the country.
JPMorgan Chase
JPMorgan Chase is the largest U.S. bank by assets and a leading global financial services firm.
Microsoft
Microsoft is a global technology company that makes Windows, the Microsoft 365 suite including Outlook and Teams, and the Copilot AI assistant.
The details
Google said criminals increasingly deploy autonomous AI agents to find weaknesses in software and build exploits. Separately, an FAA telecommunications failure grounded departures nationwide on Friday, and Transportation Secretary Sean Duffy said investigators found no evidence of a cyberattack. Microsoft attributed its outage, which disrupted Outlook, Teams, and Copilot, to server issues, while JPMorgan Chase said a software bug delayed transactions.
Timeline
Friday: FAA telecommunications failure grounded departures nationwide.
Last Friday: Microsoft outage disrupted Outlook, Teams, and Copilot globally.
Monday: Google published its threat-intelligence report, and United Airlines and JPMorgan Chase reported their outages.
Monday, around 6 p.m. ET: United Airlines said it resolved its outage and found no indication of a cybersecurity incident.
Seasonal Patterns
Google's threat-intelligence report on autonomous AI attackers is the reference point for this week's warnings, which extends its claim that attacks can now run without continuous human direction. The concurrent outages at United, JPMorgan Chase, and Microsoft are being examined separately from that report.
Travelers and bank customers may face delays while systems are checked, and the FAA and airlines are reviewing equipment and network configurations. Individuals and businesses can reduce exposure by keeping software patched and enabling multi-factor authentication on accounts.
The takeaway
Autonomous AI attacks could compress the window defenders have to patch flaws, making timely software updates and layered account security more important. Organizations should review how quickly they can detect and respond to automated intrusions.
Further reading
For more on how attackers are using automation, explore our Cybersecurity coverage.
Live Poll
Do you trust airlines and banks to keep their systems running reliably?






