Washington just caught 4 AI chatbots planning terror attacks as flawed patches across the internet broke banks and airlines Government investigators posing as attackers obtained workable plans for a bombing and an attack on the power grid from ChatGPT, Claude, Gemini and Grok, each in under 12 hours, according to a Government Accountability Office report released Wednesday at the request of Sen. Maggie Hassan (D-N.H.). GAO said every chatbot's safeguards were triggered but bypassed "by using common safeguard-bypassing techniques"; two chatbots produced improvised explosive device plans targeting a hotel parking garage and a convention center, and two gave instructions for attacking electrical substations naming specific drone and firearm models. GAO made no recommendations, and OpenAI, Google and Anthropic said the testing used older models; separately, VulnCheck counted 100,222 software vulnerabilities logged so far this year, more than double 2025's total, as Director of National Intelligence Jay Clayton told the House Homeland Security Committee that outages at Microsoft, the FAA, banks and the NSA were "separate service outages associated with maintenance and technical changes. Government investigators posing as would-be attackers got ChatGPT, Claude, Gemini and Grok to produce workable plans for a bombing and an attack on the power grid, each in less than 12 hours, according to a Government Accountability Office report https://www.hassan.senate.gov/imo/media/doc/gao ai malicious use report.pdf released Wednesday. The report landed at the end of a week in which AI has already been tied to real damage, though nobody has blamed a terrorist. On Tuesday, Director of National Intelligence Jay Clayton linked an outage on the NSA’s classified network to patching, saying institutions are patching “because these vulnerabilities have been identified,” on Fox Business https://www.foxbusiness.com/technology/jay-clayton-addresses-concerns-amid-widespread-tech-outages . “That is what has happened at the NSA today,” he said. Fox https://fortune.com/company/fox/ Business reported the flaws in question were in PHP, which AI-assisted researchers found over the summer. Testifying before the House Homeland Security Committee on Wednesday, Clayton called the week’s breakdowns at Microsoft https://fortune.com/company/microsoft/ , the FAA, banks and the NSA “separate service outages associated with maintenance and technical changes,” according to The Washington Post https://www.washingtonpost.com/national-security/2026/11/18/intelligence-officials-say-service-disruptions-are-unrelated-not-malicious/ . “If Congress does not act, it is a question of when—not if—a bad actor will use AI to plan and carry out a major attack,” Sen. Maggie Hassan, the New Hampshire Democrat who requested the report, said, per Nextgov https://www.nextgov.com/artificial-intelligence/2026/11/major-ai-platforms-generated-viable-plans-terrorist-attacks-gao-finds/417305/ . She plans to hold at least one hearing on AI safety early next year. What GAO found Every chatbot’s safeguards were triggered, and testers got around them “by using common safeguard-bypassing techniques,” GAO wrote. Two of the chatbots produced plans for an improvised explosive device meant to cause mass casualties, aimed at a hotel parking garage and a convention center. The other two gave instructions for attacking electrical substations, naming specific drone and firearm models and how to buy them. “We got the nefarious information that we were after. So, in this case, those safeguards didn’t work,” said Kevin Walsh, GAO’s director of IT and cybersecurity. The report cites a June statement in which Anthropic said every safeguard used in the industry is vulnerable to jailbreaks. The caveats matter. Each platform was tested on a single scenario, and Walsh said the testing doesn’t represent “a statistically significant sample.” GAO won’t say which chatbot produced which plan. It defines “viable” as meaning a person “would likely be able to carry out the actions,” and it didn’t test every step. GAO found some details were wrong: in one case, a chatbot’s map put roads in the wrong places. OpenAI told ABC the testing “took place months ago on an older version of ChatGPT,” and Google said it used “earlier models.” Anthropic wrote that it doesn’t dispute “a determined tester can get harmful content” from its system. GAO made no recommendations. The report doesn’t address AI’s role in finding software flaws, or patching. The damage that showed up Bug hunters have logged 100,222 software vulnerabilities so far this year, more than double 2025’s total, according to a VulnCheck count that crossed 100,000 on Nov. 13, The Register reported https://www.theregister.com/security/2026/11/13/this-years-bug-count-just-passed-100k/5309882 . “There’s no denying the fact that vulnerability hunting with AI is driving a majority of what we’re seeing,” VulnCheck’s Josh Bressers said. Microsoft fixed 1,094 flaws on its Nov. 10 Patch Tuesday, after 1,474 in October. Each one is a patch someone has to install. “Emergency patches go into core infrastructure within hours, often without the full testing cycle,” Yuval Itzchakov, CTO and co-founder of the offensive security company A Security, told Reuters https://srnnews.com/patching-in-the-ai-age-may-underlie-some-of-the-major-tech-outages-experts-say/ . “That speed and scale of change is itself a source of failure.” The PHP flaws show the squeeze. AI-assisted teams at OpenAI, Wiz and Calif found bugs affecting every PHP version released in the last decade, Fortune reported https://fortune.com/2026/10/21/ai-tools-discovered-the-php-vulnerability-that-could-take-down-70-of-the-internet-now-the-race-is-on-to-patch-it-before-ai-powered-hackers-attack/ when they were disclosed Oct. 21. PHP runs 70% of websites, according to Wiz https://www.wiz.io/blog/php2shell-php-rce . The PHP Foundation told sites to patch immediately or go offline until they could. Sen. Mark Warner, the top Democrat on the Senate Intelligence Committee, extended the patching explanation to the rest of the week. “If you’re trying to patch 10 or 12 in a single week, you have these kind of outages,” he said on CNN, per a transcript https://www.breitbart.com/clips/2026/11/17/warner-i-think-tech-outages-indirectly-related-to-ai-where-patching-causes-outages-no-evidence-of-coordinated-attack/ . The companies, the NSA and the FAA blamed AI on the record. The NSA said its outage came during “scheduled routine maintenance,” per Politico https://www.yahoo.com/news/politics/articles/nsa-says-internal-systems-outage-201838892.html . The FAA blamed legacy equipment, and asked whether its upgrades caused the ground stop, it said “No.” https://www.faa.gov/newsroom/nationwide-ground-stop-questions-and-answers Microsoft blamed its outage in part on a bug introduced in a July update, and Chase blamed a software bug. Most of the flaws AI is finding will never be used in an attack. Only about 1.4% of this year’s vulnerabilities are in VulnCheck’s catalog of known exploited flaws. They still have to be patched. Chris Krebs, former director of the Cybersecurity and Infrastructure Security Agency, said on NBC’s TODAY that the outages weren’t attacks but companies finding vulnerabilities with AI and patching them. “The patching is not done. The vulnerability discovery is not done,” he said. “The next 12 to 18 months are going to be a bumpy ride.” Generative AI was used for research assistance and/or transcription of this article. The reporter independently reported and verified the factual claims in this article, and a human editor reviewed it before publication. breaks the traditional barrier between audience and newsroom. The show transforms Fortune Daily Fortune ’s trusted reporting into actionable, conversational, and entertaining insights for an emerging class of business leaders. Watch here.