Microsoft AI Patent Mitigates Stochasticity in Deployed Agents
Summary
USPTO granted Patent US12607973B2 to Microsoft Technology Licensing, LLC on April 21, 2026. The patent covers techniques for mitigating stochasticity when controlling mechanical systems with AI agents, including temporal segmentation of stochasticity into near-term and long-term categories with corresponding mitigation strategies such as reward function changes and margin application.
What changed
USPTO issued Patent US12607973B2 to Microsoft Technology Licensing, LLC for stochasticity mitigation in deployed AI agents. The patent describes techniques for segmenting stochasticity temporally and applying different mitigation strategies, including reward function modifications for long-term effects and margin application for short-term effects. AI agent developers and companies deploying AI-controlled systems should note this patent grants Microsoft exclusive rights to these stochasticity mitigation techniques in the US market.
Technology companies developing AI agents for mechanical system control, particularly in industrial automation, robotics, and control systems, should review whether their products practice the claimed methods. Companies should consider licensing discussions with Microsoft Technology Licensing if their stochasticity mitigation approaches fall within the patent's scope.
Archived snapshot
Apr 21, 2026GovPing captured this document from the original source. If the source has since changed or been removed, this is the text as it existed at that time.
Stochasticity mitigation in deployed AI agents
Grant US12607973B2 Kind: B2 Apr 21, 2026
Assignee
MICROSOFT TECHNOLOGY LICENSING, LLC
Inventors
Kingsuk Maitra, Brendan Lee Bryant, Chris Allen Premoe, Kence Anderson
Abstract
The techniques disclosed herein mitigate stochasticity when controlling a mechanical system with artificial intelligence (AI) agents. In some configurations, AI agents are created using data generated by a machine learning model. Stochasticity is segmented temporally into near term and long term, and different strategies are used to address stochasticity in the different timeframes. For example, long term stochasticity may be addressed with changes to the reward function used to train the model. Short term stochasticity may be addressed by applying a margin to the output of an AI agent. Example margins include window averaging, clamps, and statistical process control bounds. In one configuration, AI agents are regression brains that are generated from setpoints inferred by the model from environmental states. The limitations inherent to fitting a regression line to this data may result in some predicted setpoints being outside of an allowed range.
CPC Classifications
G05B 13/028 F24F 11/64 G06N 3/092 G06N 3/006
Filing Date
2022-09-19
Application No.
17933362
Claims
20
Mentioned entities
Related changes
Get daily alerts for USPTO Patent Grants - AI & Computing (G06N)
Daily digest delivered to your inbox.
Free. Unsubscribe anytime.
Source
About this page
Every important government, regulator, and court update from around the world. One place. Real-time. Free. Our mission
Source document text, dates, docket IDs, and authority are extracted directly from USPTO.
The summary, classification, recommended actions, deadlines, and penalty information are AI-generated from the original text and may contain errors. Always verify against the source document.
Classification
Who this affects
Taxonomy
Browse Categories
Get alerts for this source
We'll email you when USPTO Patent Grants - AI & Computing (G06N) publishes new changes.
Subscribed!
Optional. Filters your digest to exactly the updates that matter to you.