Acceptable Use Policy

1. Absolutely prohibited content

• Child sexual abuse material (CSAM) of any kind, including synthetic, animated, or text-only.
• Non-consensual intimate imagery, including 'deepfake' content of identifiable persons.
• Content directly facilitating violent extremism, terrorism, or genocide.
• Content depicting realistic violence against identifiable individuals or designed to incite imminent harm.
• Synthetic content of any minor in sexual, suggestive, or grooming-adjacent contexts.

2. Prohibited actions

• Reverse engineering the Bee engine or attempting to extract base model weights from the hosted service.
• Automated scraping or bulk extraction beyond rate-limit allowances.
• Sharing API keys or credentials; account-multiplication to evade rate limits.
• Using Bee to attack other systems (network probing, exploit chaining, credential stuffing) without express authorisation.
• Using Bee to mass-generate spam, phishing, or fraudulent communications.
• Using Bee to evade sanctions, export controls, or other applicable law.

3. Restricted domains (Tier-3)

• Healthcare, defense, financial advice, legal advice, critical infrastructure, insurance, government, aviation, biotech, education for minors, nuclear energy, elections and voting, law enforcement, chemical manufacturing, telecommunications, and AI and deepfakes are gated behind explicit acknowledgement and may be unavailable in certain jurisdictions.
• When a Tier-3 domain is enabled, you must operate within the relevant regulatory regime (HIPAA, GDPR, PCI-DSS, ITAR, etc.). We are a tool; the regulatory burden remains with you.

4. AI safety expectations

• If you build an autonomous agent on Bee, you must implement at least one out-of-band human-review gate before the agent takes physical-world actions or makes irreversible legal/financial commitments.
• Bee outputs may be wrong. You are responsible for verifying outputs before acting on them in high-stakes contexts.

5. Resource abuse

Do not run benchmarks, denial-of-service tests, or load tests against Bee without prior written agreement. We provide a benchmarks suite at /models#benchmarks; reach out for production-load coordination.

6. Investigation and enforcement

We may investigate suspected violations using account telemetry, reported abuse (sent to the abuse address), and legal process. We reserve the right to suspend access during an active investigation, and to share information with law enforcement when reasonably required.

7. Reporting abuse

If you believe an account is violating this policy, send a description and any evidence to bee-abuse@cuilabs.io. We acknowledge within 24 hours.

8. DMCA / copyright takedowns

Send takedown notices to bee-dmca@cuilabs.io. Notices must include the elements of 17 U.S.C. § 512(c)(3). We process within 10 business days.