AI Transparency Notice

Last updated: May 2026

This page explains how Driven Success LLC, doing business as OpenClaw Direct (operator of openclaw.direct), approaches transparency for the AI systems hosted on our infrastructure. It applies to all Customers who deploy a Machine Instance through our Service and to End Users who interact with those Machine Instances through messaging channels such as Slack, Telegram, WhatsApp, or Discord.

Our goal is to give Customers, End Users, and regulators a clear picture of what OpenClaw Direct does, what it does not do, and how responsibilities are divided between OpenClaw Direct, the Customer, and the underlying foundation-model providers.

What OpenClaw Direct Does

OpenClaw Direct is a managed hosting provider for AI agents. We do not train, fine-tune, or develop foundation models, and we do not operate any generative AI model of our own.

Each Customer receives a dedicated, containerized Machine Instance running on our Kubernetes infrastructure in the United States. The Customer configures the Machine Instance, selects which foundation-model provider it should call, connects it to messaging channels, and defines its prompts, tools, and behavior. OpenClaw Direct operates the runtime environment and bills for the subscription, but the Customer remains the deployer and controller of the AI agent.

Foundation Models We Support

Machine Instances do not contain a built-in AI model. Instead, they call out to third-party foundation-model providers chosen and configured by the Customer. We currently support:

  • Anthropic — Claude family of models.
  • OpenAI — GPT family of models.

The Customer selects the specific model, supplies API credentials (BYOK) or uses prepaid Credits, and is responsible for complying with the chosen provider's usage policies. OpenClaw Direct does not modify model weights and has no visibility into the internal training data of these providers.

How Prompts and Outputs Flow

When an End User sends a message through a connected channel, the data follows this path:

  • End User message arrives at the messaging channel (Slack, Telegram, WhatsApp, Discord).
  • The message is delivered to the Customer's Machine Instance on our Kubernetes cluster.
  • The Machine Instance calls the foundation-model provider's API (Anthropic or OpenAI) using the Customer's selected model.
  • The provider's response is returned to the Machine Instance and forwarded back to the End User through the same channel.

OpenClaw Direct does not train any model on Customer Content. Foundation-model API keys are encrypted at rest. For full details on data categories, retention, and rights, see our Privacy Policy.

Customer Obligations as Deployer

Under the EU AI Act and similar transparency frameworks, the party that puts an AI system into use is the deployer. For Machine Instances hosted on OpenClaw Direct, that party is the Customer, not OpenClaw Direct.

As deployer, the Customer must:

  • Inform End Users that they are interacting with an AI system, unless this is obvious from context.
  • Label AI-generated content where required by law.
  • Ensure the deployment does not engage in prohibited practices.
  • Honor End User rights and complaints arising from the agent's behavior.

OpenClaw Direct supports compliance by hosting the runtime, but the Customer's prompts, tools, and channel configuration determine what is actually disclosed at runtime.

EU AI Act (Article 50)

Article 50 of Regulation (EU) 2024/1689 (the EU AI Act) becomes applicable on 2 August 2026 and imposes transparency obligations on both providers and deployers of certain AI systems, including chatbots and generative AI.

OpenClaw Direct's classification under the Act:

  • We are not a provider of a general-purpose AI model. Anthropic and OpenAI are the upstream model providers.
  • We are downstream deployer infrastructure: we host the Customer's Machine Instance, which is itself the AI system being deployed.
  • We do not permit prohibited practices under Article 5, including social scoring, real-time remote biometric identification in public spaces, and predictive policing based solely on profiling.

We commit to publishing or providing relevant technical documentation to competent authorities on lawful request.

Contact

For questions about AI transparency, model usage, or this notice, contact privacy@openclaw.direct.

For general support, including billing and Machine Instance configuration, contact support@openclaw.direct. Refunds and cancellation are addressed in our Terms §4.9.