IP Library Granted Patent US 12,423,338
Granted Patent B2
US 12,423,338 · App. 18/197,878 · Granted Sep 23, 2025

Embedded attributes for modifying behaviors of generative AI systems

Inventors: Saleema Amin Amershi (Seattle, WA); Adam Fourney (Woodinville, WA); Victor Chukwuma Dibia (Santa Clara, CA); Gagan Bansal (Seattle, WA)
Assignee: Microsoft Technology Licensing, LLC
G06F16/3329G06F9/00G06F40/20
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 12,423,338
App. No.
18/197,878
Granted
Sep 23, 2025
Kind
B2
Abstract

Systems and methods for directing behavior of a generative artificial intelligence (AI) system are provided. In particular, a computing device may obtain an input prompt associated with a requested task for one or more generative artificial intelligence (AI) systems, obtain one or more attributes based on the input prompt, modify the input prompt based on the one or more embedded attributes, and provide the modified input prompt to the one or more generative AI systems.

Claims (34)

1. A method for directing behavior of a generative artificial intelligence (AI) system, the method comprising:

obtaining an input prompt associated with a requested task for the generative AI system;

obtaining one or more embedded attributes based on the input prompt, wherein the one or more embedded attributes provide guidance for content generation by the generative AI system;

modifying the input prompt based on the one or more embedded attributes; and

providing the modified input prompt to the generative AI system.

2. The method of claim 1 , wherein obtaining one or more embedded attributes based on the input prompt includes determining if the one or more embedded attributes are embedded in one or more applications, documents, interfaces, and/or contents that are intended to communicate with the generative AI system to perform the requested task based on the input prompt.

3. The method of claim 1 , wherein the one or more embedded attributes include one or more tags, hints, markups, or indications that provide additional instructions to the generative AI system.

4. The method of claim 1 , wherein the one or more embedded attributes are embedded in an online forum or a social media site and are designed to guide a tone of a discussion or to guide a writing assistance to follow rules of the online forum or the social media site.

5. The method of claim 1 , wherein the one or more embedded attributes are different embedded attributes around each text field on a webpage and include examples of appropriate responses that can be used as guidance to the generative AI system.

6. The method of claim 1 , wherein the generative AI system includes at least one of a generative large language machine learning model, a transformer model, and/or a combination of machine learning models.

7. The method of claim 1 , wherein the one or more embedded attributes are obtained based on an association with the requested task.

8. A method for directing behavior of a generative artificial intelligence (AI) system, the method comprising:

obtaining an input prompt associated with a requested task for the generative AI system;

obtaining one or more attributes based on the input prompt, wherein the one or more attributes provide guidance for content generation by the generative AI system;

generating, in response to determining that the one or more attributes exist, a supplemental prompt based on the one or more attributes; and

providing the supplemental prompt and the input prompt to the generative AI system.

9. The method of claim 8 , wherein obtaining one or more embedded attributes based on the input prompt includes determining if the one or more embedded attributes are embedded in one or more applications, documents, interfaces, and/or contents that are intended to communicate with the generative AI system to perform the requested task based on the input prompt.

10. The method of claim 9 , wherein the one or more embedded attributes are received via one or more development tools of applications or interfaces and/or productivity tools used to generate documents or other type of content.

11. The method of claim 8 , wherein the one or more embedded attributes are embedded around the HTML textArea of the email body that provides guidance to the AI to write in the same tone as other emails previously sent to a particular recipient.

12. The method of claim 8 , wherein the one or more embedded attributes are embedded in an online forum or a social media site and are designed to guide a tone of a discussion or to guide a writing assistance to follow rules of the online forum or the social media site.

13. The method of claim 8 , wherein the one or more embedded attributes are different embedded attributes around each text field on a webpage and include examples of appropriate responses that can be used as guidance to the generative AI system.

14. A computing device for directing behavior of a generative artificial intelligence (AI) system, the computing device comprising:

a processor; and

a memory having a plurality of instructions stored thereon that, when executed by the processor, causes the computing device to:

obtain an input prompt associated with a requested task for the generative artificial intelligence (AI) system;

obtain one or more embedded attributes based on the input prompt, wherein the one or more embedded attributes provide guidance for content generation by the generative AI system;

modify the input prompt based on the one or more embedded attributes; and

provide the modified input prompt to the generative AI system.

15. The computing device of claim 14 , wherein to obtain one or more embedded attributes based on the input prompt includes to determine if the one or more embedded attributes are embedded in one or more applications, documents, interfaces, and/or contents that are intended to communicate with the generative AI system to perform the requested task based on the input prompt.

16. The computing device of claim 15 , wherein the one or more embedded attributes are received via one or more development tools of applications or interfaces and/or productivity tools used to generate documents or other type of content.

17. The computing device of claim 14 , wherein the one or more embedded attributes are embedded around the HTML textArea of the email body that provides guidance to the AI to write in the same tone as other emails previously sent to a particular recipient.

18. The computing device of claim 14 , wherein the one or more embedded attributes are embedded in an online forum or a social media site and are designed to guide a tone of a discussion or to guide a writing assistance to follow rules of the online forum or the social media site.

19. The computing device of claim 14 , wherein the one or more embedded attributes are different embedded attributes around each text field on a webpage and include examples of appropriate responses that can be used as guidance to the generative AI system.

20. The computing device of claim 14 , wherein the generative AI system includes at least one of a generative large language machine learning model, a transformer model, and/or a combination of machine learning models.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Nov 9, 2023
From: AMERSHI, SALEEMA AMIN; FOURNEY, ADAM; DIBIA, VICTOR CHUKWUMA; BANSAL, GAGAN
To: MICROSOFT TECHNOLOGY LICENSING, LLC
Reel/Frame 065512/0314 →
Continuity (1)
Related Publication 20240386038A1 · Nov 21, 2024
References Cited (28)
US 11062095B1 · Lev-Tov · 2021 [cited by examiner]
US 11468234B2 · Alba · 2022 [cited by examiner]
US 11853706B2 · Hosseini-Asl · 2023 [cited by examiner]
US 11880655B2 · Tensmeyer · 2024 [cited by examiner]
US 12111859B2 · Siebel · 2024 [cited by examiner]
US 20190163736A1 · Sharma · 2019 [cited by examiner]
US 20190213277A1 · DeLuca · 2019 [cited by examiner]
US 20220138410A1 · Meng et al. · 2022 [cited by applicant]
US 20230205980A1 · Brockett · 2023 [cited by examiner]
US 20240020538A1 · Socher · 2024 [cited by examiner]
US 20240221731A1 · Gupta · 2024 [cited by examiner]
US 20240256792A1 · Maschmeyer · 2024 [cited by examiner]
US 20240289851A1 · Williams · 2024 [cited by examiner]
US 20240320444A1 · Maschmeyer · 2024 [cited by examiner]
US 20240320476A1 · Chandrasekaran · 2024 [cited by examiner]
US 20240354309A1 · Sacheti · 2024 [cited by examiner]
US 20240354376A1 · Tandon · 2024 [cited by examiner]
US 20240354503A1 · Baruch · 2024 [cited by examiner]
US 20240354513A1 · Hong · 2024 [cited by examiner]
US 20240354555A1 · Knipfing · 2024 [cited by examiner]
US 20240362265A1 · He · 2024 [cited by examiner]
WO WO2018040341A1 · 2018 [cited by examiner]
WO WO2024220078A1 · 2024 [cited by examiner]
Liu, Zhe, et al, “Fill in the Blank: Context-Aware Automated Text Input Generation for Mobile GUI Testing”, IEEE/ACM 45th International Conference on Software Engineering, Dec. 9, 2022, pp. 1-12. (Year: 2022). [cited by examiner]
“Getting started with GitHub Copilot—GitHub Docs,” Retrieved from the Internet URL:—https://web.archive.org/web/20230428152524/https://docs.github.com/en/copilot/getting-started-with-github-copilot?tool=vscode, Apr. 28,… [cited by applicant]
Aghajanyan, et al., “HTLM: Hyper-Text Pre-Training and Prompting of Language Models,” In Repository of arxiv. 2107. 06955, Jul. 14, 2021, 14 Pages. [cited by applicant]
International Search Report and Written Opinion received for PCT Application No. PCT/US2024/025908, Jul. 4, 2024, 12 pages. [cited by applicant]
Liu, et al., “Fill in the Blank: Context-aware Automated Text Input Generation for Mobile GUI Testing,” IEEE/ACM 45th International Conference on Software Engineering, Dec. 9, 2022, 12 Pages. [cited by applicant]
Cited By (1)
US 12,608,562