skip to content
blankcollar.ai
product industries deployment workshop about
EN/DE
let's talk → login
legal

responsible ai policy.

last updated: 2026-04-19

1. why this policy exists

we build ai for european companies that need sovereignty and auditability. that creates obligations we take seriously. this policy is how we honor them in concrete product and operational terms.

2. commitments

  • sovereignty — your knowledge layer stays where you put it. swiss by default, eu on request, on-premise for regulated industries.
  • human oversight — agents suggest; people decide. write operations that alter third-party systems require explicit human approval unless you configure otherwise.
  • transparency — every answer that ships to a user includes citations. runs are logged and auditable.
  • model choice — you choose which models to route to. we do not lock you into a single provider.
  • data minimization — the agent reads what it needs, not everything. access is scoped per source per task.

3. we will not

we will not build for: weapons and military targeting systems, mass surveillance, predictive policing, biometric categorization of individuals, social scoring, automated disinformation at scale, or coercive content generation impersonating real people without consent. we will not operate a deployment that we believe violates human rights.

4. training

we do not train our own models on your content. where we route to third-party models, we use their no-training controls wherever those controls exist and are credible.

5. bias and fairness

outputs reflect the sources they are grounded in. we do not add synthetic bias. we document which sources the answer relied on. customers can audit and restrict sources that carry known bias.

6. eu ai act alignment

the product is deployed as a foundation-model-independent business tool, not as a high-risk ai system in the sense of the eu ai act. customer deployments that fall into higher-risk categories require a joint risk assessment before go-live.

7. reporting concerns

if you believe the product is being misused or misaligned with these commitments, write to hey@theblankcollar.com with the subject line "responsible ai report". we respond within 7 days.

blankcollar.ai · a product of The Blank Collar · switzerland
hey@theblankcollar.com
sovereign by design swiss hosted multi-model
privacy policy · terms of use · cookie policy · eula · disclaimer · responsible ai policy
© 2026 the blank collar. work is for bots. life is for humans.