After the 2024 U.S. presidential election, the rapid spread of misinformation online significantly affected Americans' views on candidates as well as public health, climate change, and immigration issues. The development of generative artificial intelligence, coupled with its ability to create deepfake content and its tendency to "hallucinate" facts, may further exacerbate this problem.

Robot Writing AI Writing AI Journalist

Image Source Note: Image generated by AI, image licensed from Midjourney

In response to this challenge, the Norwegian startup Factiverse was established. The company participated in the TechCrunch Disrupt Battlefield200 competition in October and won the Best Pitch Award in the categories of security, privacy, and social networks. Factiverse has developed a tool for enterprises that can perform real-time fact-checking on text, video, and audio content. The company's goal is to help businesses save research time and reduce reputational risks and legal liabilities.

Founded in 2020, Factiverse is still in its early stages and has raised approximately $1.45 million. Nevertheless, Factiverse has begun collaborating with several media and financial partners, including one of Norway's largest banks. Factiverse's CEO and co-founder Maria Amelie stated that they even provided real-time fact-checking services for the U.S. presidential debates, which were utilized by multiple media outlets.

Amelie emphasized that Factiverse is not a large language model (LLM), but rather a model built on information retrieval. As a former tech journalist and published author, Amelie has accumulated extensive experience in fact-checking in collaboration with co-founder and CTO Veena Setty. Their AI model is trained on high-quality data from reliable global sources and fact-checkers, rather than the "junk food data" used by generative AI.

The model employs machine learning and natural language processing techniques to identify claims in information in real-time and search the web, including search engines like Google and Bing, as well as academic papers. Amelie stated, "The most interesting part is that we do not simply show the first result from search engines, but rather suggest the most reliable sources to ensure the authority of the content."

Currently, Factiverse claims that its ability to identify fact-check-worthy claims surpasses that of GPT-4, Mistral7-b, and GPT-3 in 114 languages. Amelie mentioned that Factiverse has a success rate of about 80%, with future goals to attract more clients globally and improve accuracy.

In conclusion, Amelie stated, "We have enough funding to be the best, but our goal in the U.S. is to become the fastest-growing company." She also mentioned that Factiverse plans to seek another round of funding in 2025, hoping to attract clients and investors willing to invest in trust and credibility.

Key Points:

🌐 Factiverse is a Norwegian startup focused on using AI technology for real-time fact-checking, aiming to combat online misinformation.  

💼 The company has partnered with several media and financial institutions and provided fact-checking services during the U.S. presidential debates.  

📈 Factiverse claims its model has a success rate of up to 80% in identifying and verifying information, with plans to further expand its market in the future.