# US 2026/0073394 A1 · Decentralized Information-Security Processes and Systems for Dynamic, Artificial-Intelligence Centric, Auto-Generative Protection Against Cross-Channel Threat in Multi-Modal Networks

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2026/0073394 A1
- Document ID: US20260073394A1
- Status: Published application
- Kind: A1
- Application number: 18/215,274
- Filed: 2023-06-28
- Published: 2026-03-12
- Earliest claimed: 2023-06-28
- Assignees: Bank of America Corporation
- Applicants: Bank of America Corporation
- Inventors: Shailendra Singh; Sneha Padiyar
- Web page: https://patentgrep.com/patent/US20260073394A1

## CPC

- G06Q 20/4016: {involving fraud or risk level assessment in transaction processing}
- G06N 3/045: Combinations of networks
- G06N 3/08: Learning methods
- G06N 10/20: Models of quantum computing, e.g. quantum circuits or universal quantum computers
- G06N 10/60: Quantum algorithms, e.g. based on quantum optimisation, quantum Fourier or Hadamard transforms
- G06Q 20/0655: {e-cash managed centrally}
- G06Q: INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

Decentralized information-security (IS) mitigates against cross-channel threats vectors with a photonic quantum computing machine (PQCM). Threat signals are communicated to consortium members. Live model libraries regarding the threats, characteristics, metadata, model solutions, etc. are in distributed ledgers. PQCM extracts metadata and analyzes permutations to identify configuration(s) with the highest propensity to mitigate the threat. PQCM determines optimized set(s) of the threat-vector mitigation models for the configuration(s). PQCM auto-generates, dynamically by AI/ML based on the optimized set, IS rules for the configuration. Updated threats, threat characteristics, model configurations, IS rules, etc. for the cross-channel threat can be stored in distributed ledger blockchains, shared with consortium members, and deployed to prevent the threat. Nodal consensus algorithms may be used to reach agreement amongst consortium members to independently confirm threat signal validity, best model combinations, highest propensity scores, and proposed dynamically generated IS rules to address the threat.

## Claims

### Claim 1

A decentralized information-security process for preventing cross-channel threats comprising the steps of:
detecting, by a photonic quantum computing machine (PQCM), a cross-channel threat vector;
communicating, by the PQCM to a consortium of independent nodes, the cross-channel threat vector;
generating, by a consensus algorithm amongst the PQCM and the consortium, agreement on the cross-channel threat vector;
extracting, by the PQCM from a live model library, metadata for threat-vector mitigation models;
analyzing, by the PQCM based on the metadata, permutations of the threat-vector mitigation models to identify a model configuration with a highest propensity to prevent the cross-channel threat vector;
selecting, by the PQCM from the live model library, an optimized set of the threat-vector mitigation models corresponding to the model configuration;
auto-generating, dynamically by artificial intelligence on the PQCM based on the optimized set, information-security rules for the model configuration;
storing, by the PQCM in a distributed ledger blockchain, the information-security rules for the cross-channel threat vector; and
deploying, by the PQCM, the information-security rules to prevent the cross-channel threat vector.

### Claim 2 (depends on 1)

The decentralized information-security process of claim 1 further comprising the step of identifying, by the PQCM to the consortium, the distributed ledger blockchain containing the information-security rules to prevent the cross-channel threat vector.

### Claim 3 (depends on 2)

The decentralized information-security process of claim 2 wherein the permutations are analyzed by the PQCM using deep learning.

### Claim 4 (depends on 3)

The decentralized information-security process of claim 3 wherein the deep learning is based on a knowledge graph model.

### Claim 5 (depends on 3)

The decentralized information-security process of claim 3 wherein the deep learning is based on an LSTM neural net model.

### Claim 6 (depends on 3)

The decentralized information-security process of claim 3 wherein the deep learning is based on a generative adversarial network (GAN) model.

### Claim 7 (depends on 4)

The decentralized information-security process of claim 4 wherein the live model library includes fraud models for channels including mobile, credit, debit, and online.

### Claim 8 (depends on 7)

The decentralized information-security process of claim 7 wherein the information-security rules are deployed in real-time to prevent fraud in a decentralized finance (DeFi) network.

### Claim 9 (depends on 7)

The decentralized information-security process of claim 7 wherein said steps of communicating, extracting, analyzing, selecting, auto-generating, storing, and deploying are performed in real-time, after said detecting of the cross-channel threat vector by the PQCM, to prevent fraud in a decentralized finance (DeFi) network.

### Claim 10 (depends on 8)

The decentralized information-security process of claim 8 wherein the information-security rules are deployed to payment platforms.

### Claim 11 (depends on 9)

The decentralized information-security process of claim 9 wherein the information-security rules are deployed to payment platforms.

### Claim 12

A real-time information-security process for preventing cross-channel threats in a decentralized finance (DeFi) multi-modal network comprising the steps of:
receiving, by a computing machine (CM), identification of a cross-channel threat vector for the DeFi multi-modal network;
communicating, by the CM to a consortium of independent nodes, the cross-channel threat vector;
generating, by a consensus algorithm amongst the CM and the consortium, agreement on the cross-channel threat vector;
extracting, by the CM from a live model library for channels including mobile, credit, debit, and online, metadata for threat-vector mitigation models;
analyzing, by the CM based on deep learning of the metadata, permutations of the threat-vector mitigation models to identify a model configuration with a highest propensity to prevent the cross-channel threat vector;
selecting, by the CM from the live model library, an optimized set of the threat-vector mitigation models corresponding to the model configuration;
auto-generating, dynamically by artificial intelligence on the CM based on the optimized set, information-security rules for the model configuration;
storing, by the CM in a distributed ledger blockchain, the information-security rules to prevent the cross-channel threat vector; and
deploying, by the CM, the information-security rules to prevent caused the cross-channel threat vector, whereby fraud is prevented in the DeFi multi-modal network.

### Claim 13 (depends on 12)

The real-time information-security process of claim 12 wherein the information-security rules are deployed to payment platforms.

### Claim 14 (depends on 13)

The real-time information-security process of claim 13 further comprising the step of providing, by the CM to the consortium, access information in order to access the distributed ledger blockchain containing the information-security rules to prevent the cross-channel threat vector.

### Claim 15 (depends on 14)

The real-time information-security process of claim 14 wherein the deep learning is based on a knowledge graph model.

### Claim 16 (depends on 14)

The real-time information-security process of claim 14 wherein the deep learning is based on an LSTM neural net model.

### Claim 17 (depends on 14)

The real-time information-security process of claim 14 wherein the deep learning is based on a generative adversarial network (GAN) model.

### Claim 18 (depends on 14)

The real-time information-security process of claim 14 wherein the CM is a photonic quantum computing machine.

### Claim 19 (depends on 18)

The real-time information-security process of claim 18 wherein the photonic quantum computing machine analyzes all possible permutations of the threat-vector mitigation models to identify the model configuration with the highest propensity to prevent the cross-channel threat vector.

### Claim 20

A real-time information-security system for preventing cross-channel threats comprising:
a threat-vector identification platform for identifying a fraud threat in a decentralized finance (DeFi) multi-modal network including at least a mobile channel and an online channel;
a live model library containing threat-vector models to attempt to prevent the fraud threat, said live model library coupled to the threat-vector identification platform;
a photonic quantum computing machine (PQCM) to detect changes in the live model library, extract metadata from the live model library for the threat-vector models, analyze the metadata by deep learning, and compute permutations of the threat-vector models to generate a model configuration with a highest propensity to prevent the fraud threat;
a consensus module used by the PQCM to achieve consensus amongst consortium nodes on the model configuration for the model configuration most likely to prevent the fraud threat;
a blockchain retrieval module for the PQCM to extract rule sets in a distributed ledger blockchain corresponding to the threat-vector models for the model configuration;
a dynamic rule-generation module for the PQCM to dynamically generate fraud control rules derived from the rule sets extracted from the distributed ledger blockchain by the blockchain retrieval module;
a blockchain storage module for the PQCM to store, in the distributed ledger blockchain, the fraud control rules that were dynamically generated by the dynamic rule-generation module for the model configuration;
a blockchain notification module for the PQCM to provide the consortium nodes with access information to access the fraud control rules that were stored in the distributed ledger blockchain; and
a deployment module for the PQCM to deploy the fraud control rules to payment platforms to prevent the fraud threat in the DeFi multi-modal network.
