AI & Automated Systems Terms

1. Scope

These AI and Automated Systems Terms apply when a covered Aeltrix Service includes artificial intelligence, machine learning, generative systems, classification, extraction, recommendation, ranking, anomaly detection, autonomous workflow execution, or other automated functionality ("AI Features"). They supplement the Terms of Service and Acceptable Use Policy.

2. Nature of AI Output

AI outputs may be probabilistic, incomplete, outdated, ambiguous, non-unique, or factually incorrect. Aeltrix does not warrant that an output is accurate, complete, legally sufficient, professionally appropriate, or suitable for a specific decision. Customer must independently evaluate outputs before relying on them.

3. Human Oversight

Customer must maintain meaningful human oversight where AI output could materially affect a person's legal rights, employment, housing, education, credit, insurance, healthcare, access to essential services, public benefits, safety, or other consequential interests. AI Features must not be treated as the sole authoritative decision maker in high-impact contexts unless applicable law expressly permits that use and Customer has implemented required safeguards.

4. Transparency

Where applicable law requires disclosure that a person is interacting with AI, receiving AI-generated content, or being subject to automated processing, Customer is responsible for providing the required disclosure unless the Service expressly provides that notice on Customer's behalf. Aeltrix may include system-level notices, labels, metadata, or interface disclosures to support transparency obligations.

5. Inputs and Rights

Customer represents that it has all rights, permissions, and lawful bases necessary to submit prompts, documents, datasets, personal data, copyrighted works, confidential information, and other inputs to the AI Features. Customer must not intentionally submit information whose processing is prohibited by law or contract.

6. Generated Output and Ownership

As between Customer and Aeltrix, and subject to applicable law and third-party rights, Customer may use output generated for Customer through the Service. Aeltrix does not represent that output is protectable by copyright, unique, non-infringing, or free from similarity to output provided to others. Customer is responsible for clearance before publication or commercial use where required.

7. Training and Service Improvement

Aeltrix will handle Customer Content used with AI Features in accordance with the applicable contract, Privacy Policy, and DPA. Where an enterprise order states that Customer Content will not be used to train shared models, that commitment controls. System telemetry, abuse signals, operational metadata, and de-identified or aggregated information may be used as permitted by the agreement and law.

8. Third-Party Models

AI Features may use third-party model or infrastructure providers. Where applicable, Aeltrix may route requests to those providers as subprocessors or service providers. Third-party model availability, safety controls, model behavior, context windows, rate limits, and usage restrictions may change. Enterprise data-processing relationships remain governed by the applicable DPA and subprocessor framework.

9. Prohibited Uses

• unlawful discrimination or circumvention of civil-rights protections;
• malicious impersonation, fraud, phishing, or deceptive synthetic identity use;
• generation or facilitation of malware, credential theft, ransomware, or unauthorized intrusion;
• unlawful biometric identification, emotion inference, surveillance, or profiling;
• generation or distribution of unlawful sexual exploitation material or non-consensual intimate content;
• removal of legally required professional or human review;
• autonomous use in weapons or other prohibited lethal decision systems; or
• any use prohibited by applicable law or the Acceptable Use Policy.

10. High-Risk and Regulated Uses

Customer must conduct its own legal and risk assessment before deploying AI Features in regulated or high-impact contexts. Customer is responsible for required impact assessments, notices, registrations, recordkeeping, human oversight, contestability, bias testing, accessibility, monitoring, and regulator-facing documentation unless Aeltrix expressly agrees in writing to perform a specific obligation.

11. Automated Actions

Where AI Features can send messages, modify records, trigger transactions, provision resources, change permissions, or execute workflows, Customer must configure appropriate approvals, thresholds, rollback controls, permissions, and monitoring. Customer is responsible for determining when fully automated execution is appropriate.

12. Safety Controls

Aeltrix may implement filters, model routing, rate limits, content classifiers, human review, logging, refusal behavior, or other controls intended to reduce misuse. Customer may not intentionally bypass those controls except where an authorized testing agreement permits it.

13. Monitoring and Incident Response

Customer should monitor AI-enabled workflows for drift, anomalous behavior, security events, inaccurate outputs, and unexpected automation. Material incidents involving unlawful output, compromised credentials, sensitive-data exposure, or harmful automated actions should be reported promptly to Aeltrix through the applicable security or support channel.

14. No Professional Advice

AI output does not constitute legal, tax, accounting, medical, investment, financial, engineering, compliance, or other professional advice. See the Legal & Regulatory Disclaimers.