AI & LLM Policy
How BillboardBee uses artificial intelligence and large language models (LLM) in a responsible, transparent and GDPR-compliant way.
Applies to Billboard Bee AB (reg. no. 559551-9678) · Last updated: 5 February 2026
Overview
BillboardBee uses AI technology and large language models (LLM) to enhance the experience for advertisers, media owners and media agencies. Our agent helps with everything from analyzing billboard locations and generating campaign material to matching advertisers with the right screens.
This policy describes which AI services we use, how data is handled, which principles we follow and what rights you have. This policy supplements our Privacy Policy.
How We Use AI & LLM
Our agent is powered by leading language models and is used in four main areas.
Campaign Material & Mock-ups
Our agent generates campaign proposals and visual mock-ups based on your preferences. Results are always suggestions – you have full control to edit and approve.
GPT-4o / DALL·E 3Billboard Analysis & Pricing
The agent analyzes billboard locations, traffic data and demographics to calculate reach and suggest dynamic pricing.
GPT-4oIntelligent Billboard Matching
Through a conversation, the agent understands your campaign needs and matches you with the most relevant billboard locations.
GPT-4oDescriptions & Insights
The agent generates descriptions, audience analysis and recommendations for each billboard location.
GPT-4oData Handling
We take data protection seriously. Here is an overview of how data flows through our agent systems.
| Data Type | Sent to LLM? | Retention | Purpose |
|---|---|---|---|
| Billboard addresses & metadata | Yes (anonymized) | Session lifetime | Analysis & description |
| Campaign preferences | Yes | 30 days | Matching & suggestions |
| Uploaded images | Yes (for mock-ups) | Until you delete | Campaign material |
| Personal data (name, email) | No, never | Per privacy policy | — |
| Payment data | No, never | Per privacy policy | — |
Our Principles
Six core principles guide how we develop and use agent technology.
Transparency
We always clearly label when content is agent-generated. You always know whether a human or an agent produced the information.
Data Minimization
We only send the data needed for each specific task. Personal data is anonymized or pseudonymized before reaching LLM models.
No Training on Your Data
We use OpenAI's API with opt-out from model training. Your data is never used to train or improve third-party models.
Humans Decide
The agent provides suggestions and recommendations – but all decisions about publishing, booking and payment are made by you.
EU-Based Data Processing
We use OpenAI's EU data centers (Ireland) for all agent processing. Data never leaves the EU/EEA without appropriate safeguards.
Continuous Evaluation
We regularly evaluate agent output for quality, bias and accuracy. Incorrect results are reported and addressed.
Third-Party Services
These AI/LLM providers are used in our platform.
| Provider | Model | Usage | Data Center | Training on our data? |
|---|---|---|---|---|
| OpenAI | GPT-4o | Text, analysis, matching | EU (Ireland) | No (opt-out) |
| OpenAI | DALL·E 3 | Image generation | EU (Ireland) | No (opt-out) |
Frequently Asked Questions
Questions about our AI policy?
We are transparent about how we use agent technology. Have questions, feedback or want to exercise your rights? Contact us.