Iris.ai, a Norwegian AI company led by CEO and co-founder Anita Schjøll Abildgaard, builds software that reads scientific papers, patents and internal documents and extracts insights for corporate R&D teams. Its EIC application promised a "Digital R&D team member" that could "read data, analyze content and develop insights autonomously", to cut the 30% of researchers' time it said is wasted on searching and cleaning data, according to CORDIS.

The European Innovation Council granted €2,387,000 towards a €3.44 million project from May 2023 to April 2025. The final report says Iris.ai turned "fragmented TRL6 tools" into "a coherent TRL8 platform", the Researcher Workspace, "now deployed in four pilot organizations" in pharma, medtech and materials science, and that it reduces data-gathering time by up to 40%. In May 2024 Iris.ai raised a €7.64 million Series A led by Silverline Capital and matched by the EIC Fund. Its 2025 accounts, filed with Norway's Brønnøysund Register Centre, show revenue of NOK 9.6 million, a net loss of NOK 26.2 million and negative equity of NOK 47.7 million.

€2.39MEIC grant, 69% of a €3.44M project
4Pilot clients for the Researcher Workspace
NOK 9.6MRevenue, 2025
NOK -47.7MEquity, end of 2025

Company file

Legal entity
Iris.ai AS, org. no. 916 246 269, Norway
Registered
November 2015
CEO
Anita Schjøll Abildgaard, co-founder
Product
Researcher Workspace (RSpace)
Investors
Silverline Capital, EIC Fund
Total funding
About €16M by May 2024, per company
EIC project
1 May 2023 to 30 Apr 2025, closed
EIC funding
Blended finance; EIC Fund co-invested

Where Iris.ai Stands Now

Iris.ai's 2025 accounts show an operating loss of NOK 20.9 million, liabilities of NOK 87.4 million against assets of NOK 39.7 million, and current assets of NOK 13.3 million, according to the Brønnøysund Register Centre. The company is not in bankruptcy or liquidation. Its blog now focuses on enterprise AI governance, agentic systems and the EU AI Act. The company says the platform is in use via API-based enterprise deployments.

"We've onboarded a range of new clients successfully using our RSpace toolsuite for research in diverse areas like material science, food science, medical, biotech, agriculture, chemistry, automotive, and cosmetics."

Anita Schjøll Abildgaard, CEO and co-founder, Iris.ai, May 2024

The Technology: Reading Science at Scale

More than 6,000 papers and 8,000 patents are published every day, the report says. Iris.ai combines its own scientific language models with a generative AI module to summarise, filter and extract facts, and says its approach limits hallucinations. The platform can be installed on a client's own servers so confidential research data stays private.

What the EIC Project Promised, and What CORDIS Shows

The project promised a secure, scalable digital research assistant. The report documents a TRL 8 platform, four pilot deployments, 18 approved deliverables and a revised go-to-market strategy. Much of the listed impact is dissemination: conferences, webinars, social media impressions and media coverage. It does not report paying customers or revenue. CORDIS lists the project as closed.

  1. TRL 8 platformDeliveredResearcher Workspace.
  2. Pilot deploymentsDeliveredFour organisations.
  3. Follow-on fundingDelivered€7.64M Series A.
  4. Commercial scaleNot yetNOK 9.6M revenue, 2025.

The Money

The Series A was matched by the EIC Fund, which had backed Iris.ai since the January 2023 EIC selection, and took the company's total funding to about €16 million, it said. The 2025 accounts show a net loss of NOK 26.2 million and negative equity; how those losses will be funded is not stated publicly. The EIC Fund's stake is not public.

What Cannot Be Checked

Questions for the EIC

How much the EIC Fund has invested, how many paying clients Iris.ai has beyond the four pilots, and how the company plans to cover its losses are not public.

Why Europe Should Care

Europe produces a large share of the world's science but struggles to commercialise it, the application argues. A European tool that helps companies use research faster fits that gap, and Iris.ai stresses privacy and explainability as selling points. The EIC backed it with grant and equity; its accounts show the market has not yet paid off.

What the project promised

A secure digital R&D assistant that reads and analyses research autonomously, ready to scale.

What the record shows

A TRL 8 platform with four pilots, a €7.64 million Series A, NOK 9.6 million revenue and negative equity.

The Verdict

Product built, business fragile

Iris.ai delivered the platform the EIC funded. Its accounts show a company still far from covering its costs, with equity below zero.

Iris.ai and the EIC have not been asked for comment for this article. It is based entirely on public records and published statements, linked throughout.