AI Governance Practices

Streetlamp is committed to the responsible, transparent, and secure development of artificial intelligence as part of its services ("AI Services"). This page outlines our approach to AI governance and compliance.

Guiding principles

We are guided by principles of transparency, consent, and real-time safety across the AI lifecycle. We disclose the AI nature of our services, give customers choice over their data, and operate a trust-and-safety layer designed to prevent harmful interactions.

Roles and responsibilities

Our role shifts with the stage of the AI supply chain. When creating and pre-training our models we act as a "controller" under privacy frameworks such as the GDPR, and as a "provider" under AI frameworks such as the EU AI Act. When customers use the services to process their own data, they act as the controller and "deployer", and we act as a "processor". Used as intended, our services are not designed to be High-Risk AI Systems under the EU AI Act.

Data handling

Inputs to the AI Services are processed on a limited-retention basis — held only for the period needed to generate a result — unless otherwise agreed. We do not use customer inputs to train our models without permission.

Acceptable use and safety

Use of the AI Services is subject to our Acceptable Use Policy, which prohibits deceptive, abusive, or unlawful uses. Content-moderation and monitoring measures support responsible use in real time.

Explainability

Our services are designed to be understandable: predictions are accompanied, where applicable, by dimension-level scores and plain-language reasoning, so that results can be reviewed and challenged.

Fairness and bias

We apply bias-mitigation measures across the AI lifecycle, from training-data selection through monitoring, and conduct periodic assessments designed to avoid discrimination based on protected characteristics.

Governance and oversight

Defined internal roles review AI-related decisions that carry ethical or safety implications, oversee security and compliance, and govern the development and monitoring of our models.

Contact

Questions about our AI governance practices can be sent to info@streetlamp.ai.