Acceptable Use Policy

1. Purpose of This Policy

TaylinAI provides shared infrastructure for AI orchestration. To ensure a consistently reliable experience for all customers, and to maintain the security and integrity of the platform, we establish fair use guidelines covering:

• AI model token consumption (platform-brokered requests)
• API rate limits and concurrent execution
• Data storage and retention
• Prohibited AI use cases
• Responsible AI practices

2. Platform Token Usage

When you use platform-brokered LLM access (i.e. you have not configured your own API keys via Bring Your Own Key), usage is counted against your workspace's monthly token budget.

An alert email is sent to the workspace owner when 80% of the monthly budget is consumed. At 100%, platform-brokered executions return a 429 response. Budgets reset on the first day of each calendar month. Token top-up packs are available for purchase via the Billing page.

Bring Your Own Key (BYOK): Workspaces that configure their own LLM provider API keys are not subject to the platform token budget. Usage against your own keys is unlimited from our perspective; you manage limits with your provider directly.

3. API Rate Limits

The following rate limits apply to API calls made to the TaylinAI platform API:

Requests exceeding these limits receive a 429 Too Many Requests response with aRetry-After header. Persistent or automated abuse of rate limits may result in temporary IP bans or account suspension.

4. Data Storage

Fair use storage expectations per workspace:

• Prompts and versions: Up to 10,000 prompt versions (combined)
• Agents: Up to 500 agents per workspace
• Skills: Up to 200 skills per workspace
• Data connections and documents: Up to 10 GB of indexed documents per workspace
• Execution logs: Retained per your configured retention window (default 90 days)

These are fair use expectations, not hard system limits. If your usage significantly exceeds these figures and impacts platform performance for other customers, we will contact you to discuss an appropriate Enterprise plan.

5. Prohibited AI Use Cases

TaylinAI may not be used to generate, process, or distribute content or instructions that:

• Are designed to deceive or defraud individuals or organisations
• Facilitate illegal activities, including hacking, phishing, or fraud
• Generate deepfakes, synthetic media, or impersonations intended to deceive
• Produce or distribute CSAM or any material that exploits or harms minors
• Harass, intimidate, or threaten individuals
• Violate privacy rights, including generating content that reveals personal data without consent
• Infringe intellectual property rights, including reproducing copyrighted works at scale
• Generate weapons instructions, malware, exploit code, or other tools of harm
• Create or assist in creating spam, unsolicited marketing, or automated disinformation campaigns
• Circumvent safety systems of LLM providers (jailbreaking)

In addition, you must not use TaylinAI to make automated high-stakes decisions affecting individuals — such as credit scoring, employment screening, or medical diagnosis — without appropriate human oversight, disclosure, and compliance with applicable AI regulation (including the EU AI Act where applicable).

6. Responsible AI Practices

We encourage all customers to follow responsible AI practices including:

Human Oversight

Use TaylinAI's built-in governance workflows to ensure human review before AI outputs are used in consequential decisions. Do not fully automate processes where errors could cause significant harm.

Data Minimisation

Do not include unnecessary personal data in prompts sent to LLM providers. Where possible, pseudonymise or anonymise sensitive data before it enters a prompt.

Accuracy and Verification

AI-generated outputs are probabilistic and may contain errors, hallucinations, or outdated information. You are responsible for verifying all outputs before acting on them, particularly in regulated domains such as legal, medical, or financial advice.

Transparency

Where AI is used to produce communications to your end users (emails, reports, documents), we recommend disclosing that AI assistance was used, in accordance with applicable regulations and best practice.

Model Selection

Choose models appropriate for your use case and data sensitivity. Higher-sensitivity data should use providers with appropriate data processing agreements in your jurisdiction.

7. Scheduled and Automated Runs

Scheduled agent runs and automated workflows are supported. To ensure fair resource sharing:

• Schedule high-volume batch runs during off-peak hours where possible (00:00–06:00 UK time)
• Scheduled runs are subject to the same rate limits as interactive usage
• Runaway schedules (e.g. due to misconfiguration) that consume excessive resources may be automatically paused and the workspace owner notified
• Maximum concurrent scheduled agent runs per workspace: 10

8. Third-Party LLM Provider Compliance

When using platform-brokered access to LLM providers, you are also bound by the usage policies of those providers. You agree not to use TaylinAI to circumvent or violate:

• OpenAI Usage Policies
• Anthropic Acceptable Use Policy
• Microsoft Azure OpenAI Code of Conduct
• Google Gemini / Vertex AI Terms of Service
• Meta Llama acceptable use guidelines

Violation of a provider's terms may result in your access to that provider being suspended independently of your TaylinAI subscription.

9. Monitoring and Enforcement

We monitor platform usage for compliance with this policy. We do not routinely inspect the content of your prompts or outputs, but we may review usage patterns for signs of abuse (e.g. anomalous token consumption, rate limit abuse, or reports of harmful content generation).

Depending on the severity of a violation, we may:

• Issue a written warning to the workspace owner
• Temporarily throttle or suspend specific features
• Suspend the workspace pending investigation
• Terminate the subscription without refund for serious or repeated violations
• Report illegal activity to relevant authorities

We will always attempt to contact you before taking action, except where immediate suspension is necessary to protect the platform or other customers.

10. Reporting Misuse

If you observe use of the TaylinAI platform that appears to violate this policy, please report it to abuse@taylinai.com. We investigate all reports in confidence. Security vulnerabilities should be reported to security@taylinai.com.

11. Changes to This Policy

We may update this policy to reflect changes in platform capabilities, LLM provider requirements, or applicable law. Material changes will be communicated with at least 14 days' notice via email. Continued use of the platform after the effective date constitutes acceptance.