A fact-checking startup has accelerated its plans to banish AI hallucinations after securing €1 million in funding.
The Norwegian company Factiverse uses machine learning to verify content generated by artificial intelligence. The company's tools automatically detect erroneous results. These errors are now often the cause of controversy.
For example, a New York lawyer had to apologize for using fake court quotes from ChatGPT. CNET had to correct 41 of the 77 articles that the news portal had written using an AI tool. Microsoft's Bing AI produced numerous errors when analyzing earnings reports. The company's chatbot also claimed that it had spied on employees.
Factivverse offers a solution to these problems. Founded in 2019, the company develops patented models that analyze AI-generated content. These models are based on the research of Vinay Setty, the company's co-founder and CTO. Setty is also an associate professor of machine learning at the University of Stavanger. He has spent over a decade studying text and graph mining for fact-checking.
The <3 of EU technology
The latest gossip from the EU tech scene, a story from our wise old founder Boris and questionable AI art. Free in your inbox every week. Sign up now!
Setty's research laid the foundation for the Factiverse AI Editor. The tool, introduced last year, detects distortions and errors in texts created by ChatGPT and similar programs. It also shows reliable Sources for the findings.
Funding fact-checking
The new capital injection is intended to finance further improvements. Investors include Murshid Ali, the founder of Huddlestock and Skyfri, Johann Olav Koss, an entrepreneur and four-time Olympic speed skating champion, and Yasmin Namini, a former top manager at the New York Times. The investment firms Herfo and Valide Invest also participated.
CEO and co-founder of Factivverse, Maria Amelie, announced the increase today at the TNW conference.
“This investment will enable us to continue the development of our now crucial Fact Check solutions and enable even more companies to confidently harness the power of AI,” she said.
In addition to the funding, Factivverse has also launched a trial program for journalists. Users will receive free access to the fact-checking technology for 30 days. No payment card is required to test it.
Factivverse wants to use the new test version and the resources to expand the startup’s user base. Espen Egil Hansen, the company's CEO, sees AI hallucinations as a general problem.
“There are already cases where companies have to face legal consequences because of factual errors in their chatbots,” he said.“As AI penetrates ever more deeply into our daily lives, ensuring the accuracy of AI models is not just a matter of good business; it is the foundation upon which democratic values are built.”
