Newtral, a Spanish news company that focuses on fact-checking and data verification, has developed an AI tool to combat the fake news problem in Spain.
FactFlow AI is designed to monitor posts on selected channels and chatrooms, that have been nominated by the team at Newtral as being suspicious, and alert when disinformation starts to go viral.
After the work of FactFlow AI, fact-checkers at Newtral review the flagged content, to prevent fake news from spreading.
Marilín Gonzalo, the senior technology journalist in charge of Public Policy at Newtral, said: “We had many journalists listening in and checking the feeds, doing it manually, and it was impossible. So we started monitoring it all with artificial intelligence and that’s really helped us.”
The AI tool was officially presented to the public at the JournalismAI Festival in November 2025, after its development as part of the Polis JournalismAI Innovation Challenge.
It’s currently primarily in use on the messaging app Telegram, which is popular in Spain, with over 8.5 million users.
Newtral launched FactFlow AI at the JournalismAI Festival, with machine learning engineer Marta Martínez Mora explaining her team’s development. [Photo credit: Nieman Lab]
“People don’t trust anybody”
FactFlow AI was developed in response to the spread of disinformation on social media platforms, after Newtral analysed 3,000 social media posts following the devastating floods in Valencia in 2024 and found the spread of fake news to be unacceptable.
The fake news and disinformation problem is thought to be particularly large in Spain, with over 83% of people there feeling that false information is a current threat to democracy, according to a recent Eurobarmeter survey.
Additionally, over three quarters of those surveyed felt they frequently encountered false information online.
Fact-checkers at Newtral were particularly driven to create FactFlow AI following the spread of disinformation about the floods in Valencia in 2024. [Photo credit: David Ramos/Getty Images]
“People don’t trust anybody – they just feel so flooded with disinformation,” Marilín said. “They don’t trust any information that comes to their phone.”
However, Newtral is working to restore confidence in the news and online information – since its launch, FactFlow AI has been monitoring around 7,000 Telegram channels for disinformation, reviewing over 10 million messages in total.
FactFlow AI can currently can only process text, but there are plans to expand its capabilities to analysing photo and video-based content in the future.
Alongside this artificial intelligence tool, Newtral has developed several others, including “ClaimHunter”, which identifies sentences within politicians’ online posts and speeches that can be fact-checked, so that a human journalist can verify their claims against reputable sources.
Marilín feels confident that this AI-first approach is helping to reduce the fake news problem in Spain:
“If the liars are using social media to manipulate people and distribute disinformation, we have to be there too, and fight it with the same weapon they have – technology. It’s very difficult to measure, but I think people are more aware of it now.”