Bojan Stojkovski, editor-in-chief of the digital platform IT Logs, shares how this specialized tech media outlet from Macedonia is gradually integrating artificial intelligence into its newsroom operations. The outlet has already implemented a system for regularly releasing AI-generated news through Chat GPT. Now, the team aims to train the language model to autonomously verify and publish news.
The editorial team at IT Logs began experimenting with artificial intelligence in 2022, using ChatGPT and other tools to prepare interviews and generate story ideas. Later, they decided to test how generative models could support content production. Their first attempt was short video news segments with a virtual presenter, but the format didn’t gain much traction. Instead, publishing AI-generated text stories became a regular feature.
Today, the team is working on improving their language model so it can independently verify the accuracy of news and systematically publish articles on the website without human involvement.
Because part of the IT Logs team comes from a tech background, the newsroom views AI not as a threat but as a tool to reduce routine work. Generative systems handle short news updates, freeing up journalists to focus on analysis and more complex stories.
IT Logs uses artificial intelligence in several ways, constantly experimenting with new approaches. Previously, the platform published short videos of up to three minutes, where both the presenter’s appearance and voice were entirely generated by AI. However, the format didn’t resonate with audiences. After a year of experimenting with AI video news, the team decided to abandon the format and focus on other types of content generation. Today, the main application of AI is producing news articles about international technology events.
Bojan Stojkovski, editor-in-chief of IT Logs, explains the process: “You give large language models a topic that already exists — something that has happened in the tech world — and the AI builds a concept, gathers sources, translates, and rewrites.” This work was once done by journalists, but automation now allows the editorial team to focus on more in-depth reporting.
At this stage, the team’s main focus is figuring out how to make the most effective use of AI technology.
“We tried a concept where ChatGPT writes the story, and then Gemini or Grok reviews the entire piece to fact-check and make sure it’s accurate and free of misinformation,” explains Bojan.
The first AI-generated news pieces on IT Logs unexpectedly gained popularity. Readers engaged with them eagerly, and the traffic even surpassed articles written by human journalists. But with that success came criticism. The audience quickly noticed mistakes and inaccuracies. As the editor recalls, “Many people focused on the errors, because AI tends to make a lot of small inaccuracies — whether in the text itself or in the choice of words that don’t quite fit in our context. Those were the things readers most often pointed out and commented on.”
No one on the tech media team lost their job because of AI adoption. On the contrary, using OpenAI’s language model freed specialists from routine tasks and boosted their productivity. “If it used to take an editor half an hour to review a single article, AI can now do it in just 30 seconds,” the editor-in-chief gives an example of this efficiency.
The most ambitious project for IT Logs remains the development of autonomous AI agents capable not only of generating content but also of publishing it independently on the website. The team is still working on bringing the idea to life, but the tool is expected to significantly ease journalists’ workload and give them more time for what AI still can’t replicate: live conversations, in-depth analysis, and coverage of local stories.
