IP Library › Granted Patent US 12,724,684
Granted Patent B2
US 12,724,684 · App. 18/110,456 · Granted Sep 1, 2026

Artificial intelligence monitoring

Inventor: Andrew Southgate (Cambridge, GB)
Assignee: CrowdStrike, Inc.
G06F11/3065G06F11/327
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,724,684
App. No.
18/110,456
Granted
Sep 1, 2026
Kind
B2
Abstract

An artificial intelligence (AI) monitoring service detects, in real time or in near real time, misbehaving AI. The AI monitoring service monitors any of inputs to the AI, incoming/outgoing communications, API calls, inter-service/inter-container activities associated with the AI, and/or an output generated by the AI. Any activity conducted by, or associated with, the AI may be compared to an AI behavior profile defining permissible/impermissible activities. If any activity fails to conform to the AI behavior profile, alerts are sent and threat procedures are implemented. Very early stages of abnormal AI behavior are detected, thus quickly exposing abnormal AI behavior before the artificial intelligence can implement undesirable, or even harmful, actions.

Claims (33)

1 . A method that monitors an artificial intelligence (AI), comprising:

intercepting, by an endpoint AI sensor monitoring a host device, an AI output generated by the artificial intelligence;

quarantining, by the endpoint AI sensor, the AI output to a host memory device;

determining, by the endpoint AI sensor, that the AI output fails to conform to an AI output boundary; and

in response to the determining that the AI output fails to conform to the AI output boundary, deleting, by the endpoint AI sensor, the AI output quarantined in the host memory device.

2 . The method of claim 1 , wherein in response to the determining that the AI output fails to conform to the AI output boundary, further comprising at least one of disabling the artificial intelligence or reverting the artificial intelligence to a stored state.

3 . The method of claim 1 , further comprising receiving at least one of a packet header data associated with the artificial intelligence or a Uniform Resource Locator associated with the artificial intelligence.

4 . The method of claim 1 , wherein the receiving of the AI output further comprising receiving at least one of a network connection, a cloud service identifier, or an Internet Protocol address associated with the artificial intelligence.

5 . The method of claim 1 , further comprising receiving a service request associated with the artificial intelligence.

6 . The method of claim 1 , further comprising receiving an AI introspection datum provided by the artificial intelligence.

7 . The method of claim 1 , further comprising receiving a resource usage report associated with the artificial intelligence.

8 . The method of claim 1 , further comprising receiving an input associated with the artificial intelligence.

9 . The method of claim 1 , further comprising:

determining that the AI output conforms to the AI output boundary; and

releasing the AI output quarantined in the host memory device.

10 . A computer that monitors an artificial intelligence (AI), comprising:

at least one central processing unit; and

at least one memory device storing instructions that, when executed by the central processing unit, perform operations, the operations comprising:

receiving, by an endpoint AI endpoint sensor, an AI output generated by a host AI containerized service in a cloud-computing environment;

quarantining, by the endpoint AI sensor, the AI output generated by the host AI containerized service in a host memory device;

comparing, by the endpoint AI sensor, the AI output to an AI behavioral profile specifying an AI output boundary associated with a normal AI behavior;

determining, by the endpoint AI sensor, that the AI output fails to conform to the AI output boundary specified by the AI behavioral profile; and

in response to the AI output failing to conform to the AI output boundary specified by the AI behavioral profile, deleting, by the endpoint AI sensor, the AI output quarantined in the host memory device.

11 . The computer of claim 10 , wherein the operations further comprise at least one of disabling the artificial intelligence or reverting the artificial intelligence to a stored state.

12 . The computer of claim 10 , wherein the operations further comprise receiving at least one of a packet header data associated with the artificial intelligence or a Uniform Resource Locator associated with the artificial intelligence.

13 . The computer of claim 10 , wherein the operations further comprise receiving at least one of a network connection, a cloud service identifier, or an Internet Protocol address associated with the artificial intelligence.

14 . The computer of claim 10 , wherein the operations further comprise receiving a service request associated with the containerized service.

15 . The computer of claim 10 , wherein the operations further comprise receiving an AI introspection datum provided by the artificial intelligence.

16 . The computer of claim 10 , wherein the operations further comprise receiving a resource usage report associated with the artificial intelligence.

17 . The computer of claim 10 , wherein the operations further comprise receiving an input associated with the artificial intelligence.

18 . The computer of claim 10 , wherein the operations further comprise:

determining that the AI output conforms to the AI output boundary specified by the AI behavioral profile; and

releasing the quarantining of the AI output generated by the artificial intelligence.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Feb 16, 2023
From: SOUTHGATE, ANDREW
To: CROWDSTRIKE, INC.
Reel/Frame 062718/0743 →
Continuity (1)
Related Publication 20240281352A1 · Aug 22, 2024
References Cited (16)
US 10769570B2 · Lu · 2020 [cited by applicant]
US 11256989B2 · Dalli et al. · 2022 [cited by applicant]
US 20170201521A1 · Bruno · 2017 [cited by examiner]
US 20190318100A1 · Bhatia · 2019 [cited by examiner]
US 20220051344A1 · Dhuvur et al. · 2022 [cited by applicant]
US 20220067520A1 · Dalli et al. · 2022 [cited by applicant]
US 20220222440A1 · Rumman et al. · 2022 [cited by applicant]
US 20220335340A1 · Moustafa · 2022 [cited by examiner]
US 20230009127A1 · Boyer · 2023 [cited by applicant]
CN 109447048 · 2019 [cited by applicant]
CN 112085214 · 2020 [cited by applicant]
CN 112330475 · 2021 [cited by applicant]
CN 113033374 · 2021 [cited by applicant]
European Search Report mailed Jul. 22, 2024 for European Patent Application No. 24157071.2 , 7 pages. [cited by applicant]
Applicant's response to the Search Opinion submitted Feb. 11, 2025 for European Patent Application No. 24157071.2 , 19 pages. [cited by applicant]
EPO Communication (Examination Report) for European Patent Application No. 24157071.2, Southgate, “Artificial Intelligence Monitoring,” dated Feb. 2, 2026, 6 pages. [cited by applicant]