RE::AI is a Polish technology startup building AI software for media organisations. We place strong emphasis on safety, compliance, and transparency — developing products that meet the requirements of the EU AI Act and help newsrooms adopt AI responsibly. Our work is grounded in the belief that AI in media must be auditable, explainable, and accountable.

Orson AI is our unified ecosystem for journalists and newsrooms. It brings together Orson — an AI assistant for writing, research, and analysis — RE::SPARKS, a knowledge base and alert monitoring system, and Orson Rec, a tool for audio recording and transcription. All applications share a single workspace, a common context layer, and a cryptographic audit trail that makes every AI interaction verifiable.

AI Research is the third pillar of what we do. RE::AI runs an independent research programme on AI reliability in high-stakes environments, investigating how language models handle knowledge — what they know, what they don't, and how to tell the difference. Our work includes DKG (Dynamic Knowledge Gating), a system that routes AI decisions through verifiable, auditable pathways, and HDK (Hedera Detection Key), an audit trail that makes every AI interaction tamper-proof.