Acceptable Use Policy

This Policy applies to Safeera accounts, compute, inference, software, APIs and marketplace activity when incorporated into an accepted agreement. It protects users, equipment, third-party rights and service availability. It does not expand the scope of a purchased service or authorize access to another person’s information.

1. Lawful authority

Use only accounts, systems, data, models and software you are authorized to use. Obtain necessary licenses, permissions and notices for your workload. Follow applicable export, sanctions, privacy, intellectual-property and consumer-protection law. Do not conceal the actual customer, end user, destination or purpose to evade a control.

2. Prohibited conduct

• Creating, distributing or processing illegal material, including child sexual abuse material, or facilitating criminal exploitation or trafficking.
• Fraud, impersonation, theft of credentials, phishing, ransomware, destructive malware or unauthorized access to data or systems.
• Unauthorized surveillance, interception, biometric identification or disclosure of personal information.
• Attacks, denial of service, deliberate resource exhaustion, metering manipulation, fabricated jobs, double billing or concealment of unauthorized use.
• Bypassing tenant isolation, thermal protection, safety interlocks, rate limits, access restrictions, spending approvals or lawful location controls.
• Using physical systems to harm people or property, weaponizing a supported integration, or commanding a robot outside its approved operating limits.
• Infringing copyright, confidential information, trade secrets or model-license restrictions, including use of data without required rights.
• Threats, unlawful harassment or discrimination directed at people through the service.

3. Activities requiring prior authorization

Security research, penetration testing and malware analysis require authorization for the target and the relevant environment. Legitimate defensive research is not prohibited merely because it examines harmful software, but it must be contained and must not harm third parties.

Regulated or specially restricted workloads, including protected health information, export-controlled technical data or safety-critical physical control, require the applicable agreements and approved safeguards before use. Cryptocurrency mining, resale of raw account access and operation of externally accessible services require express permission where not already included in the purchased offering. AI inference tokens are usage units, not cryptocurrency.

4. Customer and host responsibilities

Customers must protect credentials, constrain agent authority and investigate misuse by authorized users. Hosts must preserve site and tenant protections and must not inspect or reuse another customer’s workloads. An administrator, equipment owner or lender has only its granted permissions. Do not upload private content to demonstrate a billing or technical issue when less sensitive evidence will suffice.

Customers remain responsible for validating AI outputs and for the lawful deployment of their applications. Safeera remains responsible for its own acts and agreed service duties. This Policy does not transfer every risk to the user or diminish statutory rights.

5. Enforcement and review

Safeera may investigate credible reports and request proportionate information. It may restrict the affected job, credential, application or service where needed to address a material violation or imminent risk. Where practicable, it will explain the issue and permit correction before suspension. Restrictions must be limited to their purpose, preserve safe equipment behavior and be reviewed for restoration.

An account restriction does not erase earned payouts, required refunds or lawful data-retrieval rights. Contractual charges and termination consequences remain governed by the accepted agreement. Safeera may make disclosures or reports required by law while protecting unrelated customer information.

6. Reporting and appeals

Report urgent security issues to security@safeera.ai and other misuse to legal@safeera.ai. Identify the affected account or resource, time and concern without including unnecessary sensitive data. Request review of a restriction by replying to the notice or contacting legal@safeera.ai with the relevant facts. Protected reports to regulators, law enforcement or legal advisers are not prohibited.

Safeera may update this Policy prospectively with material notice and any required acceptance. Updates cannot retroactively create a breach or change a fixed commercial obligation. Contact: Safeera, Attention: Legal, 101 Jefferson Dr., Menlo Park, CA 94025, United States.