Responsible AI | Geodd
Last Updated: 9th June, 2026

Responsible AI

This page explains Geodd’s approach to responsible AI use in connection with its Services, including AI inferencing, dedicated endpoints, dedicated GPUs, and bare metal infrastructure.

Geodd provides AI infrastructure. Customers decide what they build, deploy, submit, generate, and operate using the Services. Because of this, responsible AI is a shared responsibility between Geodd and each customer.

This page should be read together with Geodd’s Terms of Service, Acceptable Use Policy, Privacy Policy, Data Protection Addendum, Data Handling page, Security Measures page, Subprocessor List, and International Data Transfers page.

For responsible use, abuse, or security concerns, contact: [email protected].
For privacy or data protection questions, contact: [email protected].
For support, contact: [email protected].

01

Overview

Geodd’s role is to provide infrastructure for AI workloads. This may include:

  • Inferencing
  • Serverless Inferencing
  • Dedicated Inferencing
  • Dedicated GPU
  • Bare Metal Infrastructure
  • APIs, endpoints, dashboard, and related systems

Geodd does not generally control the customer’s application logic, prompts, datasets, model choices, outputs, end-user experience, or downstream use of AI-generated content.

Customers are responsible for ensuring that their use of the Services is lawful, safe, appropriate, and aligned with applicable obligations.

02

Our Responsible AI Principles

Geodd’s responsible AI approach is based on the following principles.

2.1 Customer Control

Customers should control their own workloads, data, prompts, outputs, models, and application behavior.

Geodd does not use customer API data for model training by default, and does not conduct human review of API prompts or outputs by default.

2.2 Data Minimization

Geodd designs its Services to minimize unnecessary data storage and reduce customer data exposure.

For AI model API services, API prompts, inputs, outputs, completions, request bodies, and response bodies are not stored by default.

2.3 Security

Geodd applies technical and organizational measures designed to protect systems, access, and customer data.

Customers remain responsible for securing their own accounts, API keys, credentials, workloads, applications, models, datasets, and access permissions.

2.4 Lawful and Safe Use

Customers must not use the Services for unlawful, harmful, abusive, deceptive, unsafe, or prohibited purposes.

Geodd may investigate, restrict, suspend, or terminate use that creates legal, security, abuse, operational, or safety risk.

2.5 Human Oversight

Customers should apply appropriate human review and oversight when AI outputs are used in important, sensitive, regulated, or high-impact contexts.

AI outputs may be inaccurate, incomplete, unreliable, unsafe, biased, or inappropriate for a customer’s intended use.

03

Customer Responsibility for AI Use

Customers are responsible for:

  • Selecting appropriate models and infrastructure
  • Designing and testing their applications
  • Reviewing and validating AI outputs
  • Providing required notices and disclosures
  • Applying human oversight where appropriate
  • Monitoring misuse or unsafe behavior
  • Ensuring lawful use of prompts, inputs, datasets, models, and outputs
  • Respecting intellectual property, privacy, and data protection rights
  • Preventing prohibited or abusive use by their own users and end users
  • Complying with Geodd’s Terms of Service and Acceptable Use Policy

Customers should not rely on AI outputs as the sole basis for decisions that may significantly affect a person’s rights, safety, health, finances, employment, education, access to services, legal position, or similar important interests unless appropriate safeguards, review, and legal compliance measures are in place.

04

AI Output Limitations

AI outputs may be wrong, incomplete, outdated, biased, misleading, offensive, unsafe, or unsuitable for a particular purpose.

Geodd does not guarantee that AI outputs will be:

  • Accurate
  • Complete
  • Reliable
  • Lawful
  • Safe
  • Non-infringing
  • Fair
  • Unbiased
  • Appropriate for customer use
  • Suitable for regulated or high-risk decisions

Customers are responsible for evaluating outputs before using or sharing them.

Customers should apply additional review where outputs are used for:

  • Medical, health, or wellness-related content
  • Legal, compliance, or regulatory content
  • Financial, credit, insurance, or investment content
  • Employment, housing, education, or eligibility contexts
  • Public-facing communications
  • Safety-sensitive or operational decisions
  • Any context where incorrect output could cause harm
05

Prohibited AI Uses

Customers must not use the Services to build, operate, support, generate, or distribute systems, content, or activity prohibited by Geodd’s Acceptable Use Policy.

Prohibited uses include, without limitation:

  • Illegal activity
  • Security abuse
  • Malware, phishing, ransomware, credential theft, or unauthorized access
  • Child sexual abuse material or exploitation of minors
  • Harassment, threats, hate, abuse, or doxxing
  • Fraud, impersonation, or deceptive activity
  • Spam or abusive messaging
  • Unlawful surveillance or unauthorized tracking
  • Harmful instructions for violence, terrorism, criminal harm, or self-harm
  • Non-consensual sexual content
  • Infrastructure abuse
  • Regulated or high-risk use without required safeguards and contractual permission
  • Intellectual property or data rights violations
  • Cryptocurrency mining unless expressly authorized in writing

Geodd’s Acceptable Use Policy is available at: geodd.io/legal/acceptable-use-policy

06

Regulated and High-Risk Uses

Unless expressly agreed in writing, Geodd does not provide Services specifically intended for workloads requiring dedicated contractual, regulatory, or compliance controls, including:

  • HIPAA-regulated protected health information
  • Healthcare or medical diagnosis, treatment, or clinical decision-making
  • Children’s data or services directed to children
  • Biometric identification or surveillance
  • Criminal offense records
  • Law enforcement or criminal justice decisions
  • Financial, credit, insurance, employment, housing, education, or eligibility decisions that produce legal or similarly significant effects
  • Critical infrastructure or safety-critical systems
  • Other regulated, sensitive, or high-risk uses requiring dedicated safeguards

If a customer uses the Services for any regulated, sensitive, high-risk, or legally restricted use, the customer is responsible for ensuring that it has all required lawful bases, notices, consents, authorizations, permissions, safeguards, approvals, and contractual rights.

Geodd may suspend or restrict use that creates a material legal, security, regulatory, operational, or compliance risk.

07

Human Review and Oversight

Customers should use human review where AI outputs may affect people, safety, rights, access to services, legal obligations, or business-critical decisions.

Human oversight may include:

  • Reviewing outputs before publication or action
  • Testing model behavior before deployment
  • Monitoring production use
  • Reviewing edge cases and failure modes
  • Setting confidence thresholds or escalation rules
  • Logging customer-side decisions where legally appropriate
  • Providing end users with appeal or correction processes where required
  • Avoiding fully automated high-impact decisions unless legally supported

Customers are responsible for determining the level of human review appropriate for their use case.

08

Transparency and Disclosure

Customers are responsible for deciding whether their users or end users must be told that they are interacting with AI or receiving AI-generated content.

Customers should provide clear disclosures where required by law or where failure to disclose would be misleading.

Examples may include:

  • AI chatbots
  • AI-generated recommendations
  • AI-generated summaries
  • AI-generated content shown to end users
  • Automated support tools
  • Decision-support tools
  • Public-facing AI-generated content

Customers must not represent AI outputs as human-generated where disclosure is required by law or where failure to disclose would be deceptive.

09

Data and Privacy Responsibilities

Customers are responsible for ensuring that data submitted to the Services may lawfully be processed.

Customers should avoid submitting unnecessary personal data, sensitive data, regulated data, confidential information, or high-risk data unless their use case is lawful, appropriate, and supported by the applicable agreements.

Customers are responsible for:

  • Having a lawful basis for submitting personal data
  • Providing required privacy notices
  • Obtaining required consents, authorizations, or permissions
  • Respecting data subject rights
  • Selecting appropriate infrastructure regions
  • Using appropriate safeguards for sensitive or regulated data
  • Complying with applicable data protection laws

Where Geodd processes Customer Personal Data as a processor, Geodd’s Data Protection Addendum applies.

Geodd’s Data Protection Addendum is available at: geodd.io/legal/data-protection-addendum

Geodd’s Data Handling page is available at: geodd.io/trust-center/data-handling

10

No Training Use by Default

Geodd does not use API prompts, inputs, outputs, completions, request bodies, response bodies, uploaded files, embeddings, customer datasets, or fine-tuning data for model training by default.

Geodd also does not use API data for fine-tuning by default.

If any customer-specific training, fine-tuning, evaluation, or dataset processing is offered in the future, it should be governed by a separate written agreement or Order.

11

No Human Review of API Prompts or Outputs by Default

Geodd does not conduct human review of API prompts or outputs by default.

Geodd personnel may access limited account, usage, technical, operational, billing, support, or security information where reasonably necessary to:

  • Provide support
  • Investigate technical issues
  • Debug service issues
  • Secure the Services
  • Prevent fraud or abuse
  • Comply with legal obligations
  • Enforce agreements and policies
  • Operate and maintain the Services

Such access is limited based on role, operational need, and applicable security controls.

12

Abuse Prevention and Enforcement

Geodd may investigate suspected misuse of the Services.

Where permitted by law and applicable agreements, Geodd may take action to protect customers, users, Geodd, infrastructure, networks, providers, and third parties.

Actions may include:

  • Warning the customer
  • Requesting removal or correction of prohibited activity
  • Restricting specific features, endpoints, workloads, traffic, or regions
  • Suspending access to the Services
  • Disabling API keys or credentials
  • Removing or blocking unlawful or abusive content where applicable
  • Reporting activity to law enforcement, regulators, or affected third parties where legally required or appropriate
  • Terminating Services in accordance with the applicable agreement

Geodd may take immediate action without prior notice where necessary to prevent harm, protect security, comply with law, prevent service disruption, protect other customers, or respond to abuse, fraud, or unlawful activity.

13

Dedicated GPU and Bare Metal AI Workloads

For Dedicated GPU and Bare Metal Infrastructure, customers may have direct control over workloads, models, datasets, credentials, operating environments, access permissions, and network exposure.

Customers are responsible for:

  • Securing customer-controlled systems
  • Managing deployed models and datasets
  • Ensuring model and dataset rights
  • Preventing prohibited or unlawful workloads
  • Monitoring customer-controlled deployments
  • Managing output review and downstream use
  • Applying patches where customer controls the environment
  • Maintaining independent backups where required
  • Complying with applicable law and Geodd policies

Customers must not use Dedicated GPU or Bare Metal Infrastructure to host, run, train, generate, or distribute prohibited content, malware, abuse infrastructure, unauthorized surveillance systems, or unlawful AI workloads.

14

Model, Dataset, and License Responsibility

Customers are responsible for ensuring that they have the rights required to use any models, datasets, software, prompts, outputs, and workloads they submit to, deploy on, or use with the Services.

This includes responsibility for:

  • Model licenses
  • Dataset licenses
  • Software licenses
  • Open-source license obligations
  • Third-party terms
  • Intellectual property rights
  • Privacy and data protection rights
  • Export control or sanctions restrictions
  • Use restrictions attached to third-party models or datasets

Geodd is not responsible for customer-selected models, customer datasets, customer software, customer prompts, customer outputs, or customer-controlled workloads unless expressly agreed in writing.

15

Reporting Abuse or Responsible AI Concerns

To report abuse, harmful use, or responsible AI concerns, contact: [email protected]

Reports should include, where available:

  • Description of the issue
  • Relevant URLs, endpoints, logs, timestamps, or account information
  • Evidence of abuse or harmful activity
  • Contact details for follow-up

Customers should not include API keys, passwords, secret tokens, private keys, or unnecessary sensitive data in reports.

16
17

Changes to This Page

Geodd may update this Responsible AI page from time to time to reflect changes in law, AI risks, abuse patterns, Services, infrastructure, safety practices, or operational practices.

If changes are material, Geodd will provide notice by reasonable means, such as through the website, dashboard, email, or other electronic notice.

The “Last Updated” date shows when this page was last revised.

18

Contact

For responsible AI concerns, abuse reports, or security issues: [email protected]

For privacy or data protection questions: [email protected]

For support: [email protected]

For legal notices: [email protected]

Geodd LLC

1007 N Orange St., 4th Floor, Suite #1382

United States

Website: geodd.io