Skip to main content
CitiPatent grant

Citibank was granted a patent for validating machine-learning outputs safely.

Source: Autobound Signal API · 65 signals tracked for Citi · updated

Get Citi's live signals in Claude or ChatGPT, or by API

AskWhat changed at Citi this week? Use Autobound signals for citi.com.

What happened

In October 2026 Citibank patented an isolated testing platform for checking AI-generated code and outputs before use in financial software environments.

Source

Abstract

The systems and methods disclosed herein enable evaluation of machine learning model outputs within a virtual environment. The disclosed model validation platform enables testing of code generated for detection of malicious or anomalous outputs. For example, the model validation platform can construct a virtual machine isolated from the system and test model-generated code for validation of LLM-generated outputs. In some implementations, the model validation platform determines parameters of the virtual machine and/or associated validation test based on an evaluation of the machine learning model's output and/or the associated underlying prompt. For example, the parameters of the validation test depend on an evaluation of the user or the provided input (e.g., depending on the presence of sensitive data within the prompt). By doing so, the system enables dynamic evaluation of machine learning model outputs to improve the security and robustness of associated generated code.

patents.google.com/patent/US12762040Read the full source

Extracted by Autobound

From the Signal API record
Signal
Patent grant

What this signalsA new patent often shows where a company is putting its R&D budget.

Grant date
Oct 13, 2026
Inventors
4
Tech area
software / computing
Patent number
US12762040

The full record

From the Signal API record

People

  • Payal JainInventor
  • Tariq Husayn MaonahInventor
  • Mariusz SaternusInventor
  • Daniel LewandowskiInventor
  • Biraj Krushna RathInventor
  • Stuart MurrayInventor

Details

Primary CPC class
G06F 11/3608 (Electric digital data processing)
USPTO assignee
CITIBANK. N.A.
What the invention does
A sandbox that runs machine-learning-generated code away from live systems to detect malicious or unusual behavior before trusting the result.

Topics and mentions

Tags

  • Intellectual Property
  • Research & Development
  • Artificial Intelligence
  • Machine Learning
  • Cyber Security

Technologies named

  • machine learning
  • AI validation
  • sandboxing
  • virtual machines

Extraction

Detected
Oct 14, 2026
signal_type
patents-company
signal_subtype
patentGrant

Use this data

Get every patent signal for Citi and the companies you sell to, in the tools you already use.

  1. Ask Claude about it

    Connect Autobound to Claude, Claude Code or Cursor with MCP. Then ask: “What changed at Citi this week?”

  2. Send it to your own tools

    The Signal API returns patent signals for any list of companies as JSON, for your CRM, warehouse or app.

  3. Try it free

    Sign up and spend your free credits on the companies you sell to.

    Start Free1,000 free credits

The API returns more than this page shows

This page shows a preview. The full patents-company record in the Signal API and MCP can also have these 8 fields. Some fields are empty for some signals.

Company

  • linkedin_urlValue in the API
  • industriesValue in the API
  • employee_count_lowValue in the API
  • employee_count_highValue in the API
  • revenueValue in the API
  • descriptionValue in the API

Signal

  • signal_nameValue in the API
  • associationValue in the API
Show the full JSONThe record on this page and the API request

GET /v1/signals/0ed6d196-5a24-5896-a2df-33c17b3d6a5c returns this record as JSON. POST /v1/companies/enrich returns every signal for citi.com.

{
  "signal_id": "0ed6d196-5a24-5896-a2df-33c17b3d6a5c",
  "signal_type": "patents-company",
  "signal_subtype": "patentGrant",
  "detected_at": "2026-10-14T15:49:12+00:00",
  "company": {
    "name": "Citi",
    "domain": "citi.com"
  },
  "data": {
    "cpc": [
      {
        "code": "G06F 11/3608",
        "label": "Electric digital data processing"
      },
      {
        "code": "G06F 8/41",
        "label": "Electric digital data processing"
      }
    ],
    "tags": [
      "Intellectual Property",
      "Research & Development",
      "Artificial Intelligence",
      "Machine Learning",
      "Cyber Security"
    ],
    "detail": "In October 2026 Citibank patented an isolated testing platform for checking AI-generated code and outputs before use in financial software environments.",
    "summary": "Citibank was granted a patent for validating machine-learning outputs safely.",
    "event_at": "2026-10-13",
    "inventors": [
      {
        "name": "Payal Jain",
        "status": "unresolved",
        "is_primary": true
      },
      {
        "name": "Tariq Husayn Maonah",
        "status": "unresolved",
        "is_primary": false
      },
      {
        "name": "Mariusz Saternus",
        "status": "unresolved",
        "is_primary": false
      },
      {
        "name": "Daniel Lewandowski",
        "status": "unresolved",
        "is_primary": false
      },
      {
        "name": "Biraj Krushna Rath",
        "status": "unresolved",
        "is_primary": false
      },
      {
        "name": "Stuart Murray",
        "status": "unresolved",
        "is_primary": false
      },
      {
        "name": "Philip Davies",
        "status": "unresolved",
        "is_primary": false
      }
    ],
    "tech_area": "software / computing",
    "event_kind": "grant_date",
    "grant_date": "2026-10-13",
    "source_url": "https://patents.google.com/patent/US12762040",
    "cpc_primary": {
      "code": "G06F 11/3608",
      "label": "Electric digital data processing"
    },
    "patent_title": "Dynamic input-sensitive validation of machine learning model outputs and methods and systems of the same",
    "patent_number": "US12762040",
    "uspto_assignee": "CITIBANK. N.A.",
    "patent_abstract": "The systems and methods disclosed herein enable evaluation of machine learning model outputs within a virtual environment. The disclosed model validation platform enables testing of code generated for detection of malicious or anomalous outputs. For example, the model validation platform can construct a virtual machine isolated from the system and test model-generated code for validation of LLM-generated outputs. In some implementations, the model validation platform determines parameters of the virtual machine and/or associated validation test based on an evaluation of the machine learning model's output and/or the associated underlying prompt. For example, the parameters of the validation test depend on an evaluation of the user or the provided input (e.g., depending on the presence of sensitive data within the prompt). By doing so, the system enables dynamic evaluation of machine learning model outputs to improve the security and robustness of associated generated code.",
    "precision_class": "structural",
    "invention_explanation": "A sandbox that runs machine-learning-generated code away from live systems to detect malicious or unusual behavior before trusting the result.",
    "technologies_mentioned": [
      "machine learning",
      "AI validation",
      "sandboxing",
      "virtual machines"
    ]
  }
}

Long text fields are shortened on this page.

Looking up one signal by its id is free. Enrich costs 2 credits per signal returned; a call with no results is free.