HS and IS establish a joint team to use generative artificial intelligence in newsrooms


Sustainability Sanoma Media

The joint artificial intelligence team of Helsingin Sanomat and Ilta-Sanomit start their work at the beginning of the year. In the picture, from left to right, Juho Salminen, Antti Lehtonen, Päivi Ala-Risku, Esa Mäkinen, Niina Viitanen, Pauli Tölli and Matti Järvi. Photo Saara Mansikkamäki / HS

The purpose of the HS–IS Lab launching at the beginning of the year is to test, develop and deploy generative artificial intelligence tools and combine journalistic and technological expertise. The team is also tasked with spreading understanding of artificial intelligence and its more extensive use to editorial teams. Other projects investigating the use and opportunities of artificial intelligence will continue unchanged across Sanoma.

“Generative artificial intelligence is likely to change editorial work as much as the introduction of the internet did at its time. Therefore, it is important for the editorial teams to be onboarded to this transformation at an early stage,” Johanna Lahti, the Editor-in-Chief of Ilta-Sanomat, justifies the establishment of the HS–IS Lab.

The purpose is to provide journalists with more day-to-day practical opportunities to test, brainstorm and also critically evaluate tools that have impacts on their own work.

“AI provides news desks with lots of opportunities. However, it also involves major questions relating to trust, validation of information and transparency. Therefore, it cannot be an exclusively technological question in any case,” says Erja Yläjärvi, Senior Editor-in-Chief of Helsingin Sanomat.

The HS–IS Lab engages people from the editorial teams of Helsingin Sanomat and Ilta-Sanomat as well as Sanoma’s joint news media digital development team. All of the tools and observations developed by the Lab will also be shared with the other Sanoma news media. In the initial stage, the project will employ between five and seven people on average. Pauli Tölli, who has been supervising several Sanoma artificial intelligence projects, will start as the project leader.

“AI-related technologies are evolving in great strides. It is great to be able to experiment and develop different ways of using it with the smart people of the editorial teams. I believe that together we will find the best ways for news desks to use artificial intelligence responsibly,” says Tölli, describing the opportunities.

Pauli Tölli, who has been supervising several artificial intelligence projects at Sanoma, will start as the head of the HS–IS Lab. Photo: Samuli Pulkkinen

“Sanoma has been working on artificial intelligence and, before that, machine learning for over a decade, and this work will continue at full steam across the company. Yet, journalistic work is a case of its own, with artificial intelligence perhaps involving even more ethical questions than in many other tasks,” Tölli says.

In artificial intelligence projects, Tölli balances between being a brake and inciting excess hype.

“At Sanoma, we can identify numerous blanks, but we need to come up with a smart funnel for what to investigate in more detail and what could finally end up in production,” he says.

Tölli has also considered the winners and losers of this evolution. In any case, artificial intelligence will gain ground, regardless of whether Sanoma will use it or not. News brands may well be among the winners, as the increasing use of fake news and false images will increase the need for reliable information. International technology giants may also be among the winners, but they should not be allowed to corner the entire market.


“Reliability is one of our key values. If someone wants to hurt us, they might aim to erode our trustworthiness.”
Pauli Tölli


Common people, on the other hand, easily end up on the losing side, with increased difficulty in telling right information from false.

“Reliability is one of our key values. If someone wants to hurt us, they might aim to erode our trustworthiness,” Tölli says.

“Ultimately, I believe there will be more opportunities than threats for us. Nevertheless, I don’t want to be Sanoma’s AI guru who will solve all problems. We actually have a huge number of smart people who come up with all the things where this could be used.”