Sweden’s Redpine Raises €1.1 Million To Combat AI Hallucinations With Licensed Data

The funding round included prominent angel investors and partnerships with leading foundational AI companies and content owners worldwide. Key investors include Colin M. Evans (OpenAI), Feng Hong (Xiaomi co-founder), Anna Nordell Westling (Sana co-founder), Daniel Langkilde (Kognic founder, ex-Recorded Future), Gustav Lindqvist (Perplexity), and several Spotify alumni.

“AI still operates with critical blind spots,” said Redpine’s CEO and Co-founder, Anders Hammarbäck. “We reduce hallucinations and boost accuracy by giving AI systems access to the right data – real-time, high-quality, and licensed. That’s how we help AI not just respond, but reason, decide, and act in the real world.”

Redpine’s €1.1 million funding places it at the early stage of 2025’s European AI startup landscape, compared with other infrastructure, tooling, or data platform companies already raising multi-million euro rounds. Investor appetite for AI infrastructure and data-adjacent plays remains strong, with compute platforms (Nscale, DataCrunch), model tooling (yasp), and data/test services (Synthesized) attracting larger capital.

Unlike compute or hardware-focused startups, Redpine’s licensing-driven, domain-specific data model positions it in the emerging “data as infrastructure/AI content licensing” niche. Demonstrating traction in high-risk sectors like healthcare, law, and finance, while forging strong partnerships with content owners and foundations, could open doors for follow-on funding.

Founded in 2024 by Anders Hammarbäck (ex-McKinsey, Antler) and David Österdahl (ex-Spotify, iZettle), with Leonora Vesterbacka (PhD, CERN; AI R&D at KBLab) leading data science, Redpine provides AI builders and autonomous agents with secure, licensed, high-quality multi-modal data at scale. The platform enables content owners including copyright holders and proprietary data providers to offer compliant, controlled access.

Redpine aims to solve one of AI’s biggest challenges: the shortage of trustworthy, domain-specific data in critical areas where accuracy is essential, such as healthcare, law, finance, and robotics.

“They’re solving one of AI’s biggest challenges – access to high-quality data – and doing it globally. Redpine is a standout from the new wave of Nordic AI startups. I’m proud to be an early backer,” said Colin M. Evans, OpenAI.

Redpine’s platform provides AI developers and autonomous agents with access to high-quality, multi-modal, domain-specific datasets for all stages—from pre-training and fine-tuning to inference and real-time RAG pipelines. It includes text, images, video, audio, and code across sectors like medical, legal, scientific, and software development.

The company already offers over 100 billion tokens of premium data, with a rapidly growing repository. Redpine targets the global AI training data market, an emerging segment of the trillion-dollar AI industry, expanding at 24.9% annually. Data is licensed directly from copyright holders, proprietary dataset owners, and publishers.

With increasing legal scrutiny on unauthorized AI training—highlighted by Anthropic’s $1.5B book copyright settlement and new EU disclosure rules—Redpine’s compliance-first approach provides a vital, legally sound alternative for AI development.

“For AI to reach its full potential, we need both the highest quality data available to train, develop, and fine-tune, as well as fair compensation paid to the creators and owners of such materials to ensure sustainability into the joint future. Redpine exists to provide both,” added Roy Kaufman, Managing Director, Business Development at Copyright Clearance Center

About Redpine

Founded in 2024, Stockholm-based Redpine provides AI builders and autonomous agents with secure, licensed, high-quality, multi-modal data. The platform enables content owners, including copyright holders and proprietary data providers, to offer compliant access, unlocking value for both data providers and AI developers. Backed by top investors.