Skip to main content

AI-assisted vulnerability discovery raises patching and outage concerns

22 NOVEMBER 2026·2 MIN READ·4 SOURCES

Cybersecurity experts expect outages and disruptions to continue as organizations use AI to find and fix software vulnerabilities, Bloomberg reported. A Nov. 20 report from Banking Dive says experts expect more outages as AI models expose weaknesses, including at banks.

Key takeaways · 4

  • 01

    Citi CEO Jane Fraser said patching volume was five to 10 times what the world’s technology stack had previously experienced.

  • 02

    U.S. Bank CISO Ann Barron-DiCamillo said teams should prioritize vulnerabilities that are externally exposed, exploitable and connected to critical services.

  • 03

    Mayer Brown partner Justin Herring said banks should identify critical vendors and review business-continuity plans and contract protections.

  • 04

    Barron-DiCamillo said AI is compressing the time between vulnerability discovery and exploitation, rather than necessarily creating novel attacks.

More vulnerabilities, faster decisions

Experts told Banking Dive to expect more outages as AI models expose vulnerabilities, including at banks.[2] Herring said the rate at which vulnerabilities are found has outpaced the capacity to fix them.[2] Fraser said outages and downtime would likely continue while technology is upgraded and vulnerabilities are patched.[2] She also warned that powerful AI models could pose a problem when used by bad actors.[2] For security leaders, the issue is not simply how many findings arrive, but how quickly teams can assess and address the ones that matter most.

Patching volume is not a security score

Fraser said patching volume was five to 10 times what the world’s technology stack had previously experienced.[2] Barron-DiCamillo cautioned that finding more weaknesses should not automatically be read as declining security, and said counting vulnerabilities is no longer a useful metric.[2] Her suggested priority is weaknesses that are externally exposed, exploitable and connected to critical services.[2] That approach gives teams a basis for sequencing work when they cannot address every finding at once.

Outages have varied explanations

Banking Dive reported outages at Microsoft, the FAA, JPMorgan Chase, United Airlines, Apple and the NSA over the preceding eight days.[2] Fortune reported that officials said recent nationwide outages were not the result of a coordinated cyberattack.[4] A spokesperson for Senator Mark Warner’s team told Fortune that botched patch implementation was linked to an NSA communications outage and banking outages.[4] JPMorgan Chase said OpenAI’s tools were not involved in its outage, which the bank attributed to a temporary software bug in on-premise hardware systems.[4] These accounts do not establish that one cause explains every reported outage.

Prepare for vendors and faster response

Herring said banks will depend more on outside service providers because AI tools are generally built on external models.[2] He advised banks to identify critical vendors and understand how continuity plans and contract protections mitigate vendor risk.[2] He also said banks should test incident-response plans with decision-makers and update them for faster action and new tools.[2] Barron-DiCamillo’s warning that discovery and exploitation are getting closer together reinforces the value of rehearsing decisions before an incident.[2] Teams can use these recommendations to review ownership, escalation paths and vendor dependencies.

Security and technology teams face a balancing act: patch quickly, but account for the possibility that updates can introduce failures. Prioritizing exploitable weaknesses tied to critical services and rehearsing response decisions can help leaders direct limited attention where it matters most.

Why it matters
For teams

AI training for your team

kju.ai is built for teams: daily sessions, team analytics, and rollout support for your organization.

How this developed

  1. 22 November 2026

    AI-assisted vulnerability discovery raises patching and outage concerns

Sources

AI fluency, one session a day, built for your work.