Denmark’s AI news start-up faces scrutiny over editor Michael Dyrby’s past

More on this

1. https://www.dagensmedia.se/tidningar-och-tidskrifter/dagspress/danska-ai-nyheter-praglade-av-metoo/ - Please view link - unable to able to access data
2. https://y.dk/om-y - Y’s ‘About’ page presents the Danish outlet as a news service covering politics, society, business, economics, international affairs, technology and the future. It describes the publication as politically centre-right and says it aims to be clear about that outlook while testing its own assumptions and placing factual accuracy above sharp opinion. The page explains that artificial intelligence is used to identify developments, process large amounts of information, prepare drafts and provide material in different formats. It states that editors select, assess and approve the journalism, and that nothing is published before editorial review. AI-created images, audio and video are labelled, as is synthetic-voice narration. The page also sets out the outlet’s stated commitment to editorial independence: owners, advertisers and partners do not determine which stories are covered or what conclusions are reached. Commercial content is to be clearly distinguished from editorial material. It acknowledges that errors can occur and says significant factual mistakes will be corrected promptly, with corrections made clear when they affect a story’s meaning. Michael Dyrby is identified as editor-in-chief, and the page names Projekt Y ApS as publisher. This is the outlet’s own account of its methods and editorial standards, rather than independent verification of how those standards operate in practice.
3. https://y.dk/ai-politik - Y’s artificial-intelligence policy explains how the outlet says it uses AI in news production and where human editorial responsibility remains. It says AI can monitor sources and identify developments that may become stories, process large quantities of material, draft articles based on sources, translate content and help produce different formats, including audio, summaries and video. The policy says AI does not decide what Y publishes and does not publish material independently: an editor must review and approve every item, including text, audio, video, images and live blogs. It assigns editorial and press-ethics responsibility to the editor-in-chief. The page says AI-generated or substantially altered images, audio and video are labelled, and that synthetic-voice narration is identified when playback begins. Articles do not receive an AI label, it explains, because an editor has reviewed and approved them; sources are instead disclosed. Y says it has agreements with Ritzau, Ritzau Scanpix and Reuters, with sources and photographers credited alongside content. The policy acknowledges that language models may make errors or omit context and says editors approve material before publication and corrections are made openly. It also describes how user inputs to Y’s AI features are handled and states that those inputs are not used to train language models. The policy was last updated on 27 October 2026.
4. https://projekty.dk/ - Projekt Y’s corporate website describes the venture as both a news outlet and a technology platform combining journalism, factual information and artificial intelligence. Its stated aim is to help audiences understand events and developments from fresh perspectives. The site presents the project’s values as factual documentation, transparency, responsibility, individual freedom, constructive challenge and progress. It says technology should make the world easier to understand, while human judgement remains responsible for direction. The website identifies Michael Dyrby as editor-in-chief and lists a wider team spanning editorial, product, technology, marketing, audiovisual production and AI-related roles. It introduces Y as a digital news service intended to make it easier to follow important public affairs in a format that suits each user. A second element, Y rails, is described as an AI-native media-production system intended to coordinate automated agents, workflows and modular technical infrastructure. The company presents its news outlet as the first application of that platform and suggests that the underlying system could support the creation and distribution of journalism more broadly. This source helps substantiate the article’s description of Y as a new Danish media project associated with Dyrby and AI, while also showing how the company itself characterises its ambitions. As a corporate presentation, it sets out the organisation’s own claims and objectives, not independent assessments of its journalism or technology.
5. https://www.medietrends.dk/2025/08/28/projekt-y-bliver-det-journalistik-eller-noget-der-kaldes-journalistik/ - This Medietrends commentary examines whether Projekt Y’s proposed AI-supported output should be regarded as journalism and how much human work the project will involve. Published on 28 August 2025, it reports that Michael Dyrby described the approach as journalism using AI under human direction. The article also discusses claims that the system would draw on a database of verified sources and data, and notes Supertrends’ connection to investor Lars Tvede. The author distinguishes the broad task of collecting, processing and communicating information from journalistic practices such as interviewing, source criticism, seeking responses, eyewitness reporting, transparency and building relationships with sources. In the author’s view, AI can process existing information but cannot independently conduct interviews, pursue new evidence, develop trust with sources or report directly from events. The commentary argues that an AI news operation therefore faces a choice: employ people to create the reporting and evidence that AI can help process, or rely on material gathered from existing online sources, potentially raising technical and legal concerns. It also stresses that the project’s eventual operation was not yet clear at the time of publication. This is an opinion and analysis piece, not a report of a lawsuit or a definitive finding about Y’s practices. It provides context for the debate over AI, editorial work and the meaning of journalism that surrounds the article’s account.
6. https://www.ekkofilm.dk/artikler/dyrby-tager-hjem-efter--anklage-om-sexisme/ - Filmmagasinet Ekko’s report describes the controversy surrounding Michael Dyrby after allegations of sexist and boundary-crossing conduct became public in 2021. It says Dyrby, then editor-in-chief of B.T., took several days away from work amid criticism relating to his earlier time at TV 2 and the culture depicted in a documentary about sexism at the broadcaster. The article recounts allegations made by former TV 2 employees, including an account by Anna Thygesen concerning an exchange in 2007; it also notes that Dyrby denied making the quoted remark. Ekko reports that Dyrby left his TV 2 news-director role in 2015 after compromising images circulated in connection with a personal relationship, and that he later returned to senior media leadership at B.T. The report describes a crisis meeting at the newspaper and criticism directed at Berlingske Media’s leadership for its handling of the matter. It distinguishes allegations and reported accounts from Dyrby’s own responses, including his statement that he and the group chief executive had agreed he would take a few days off. This article is relevant to the background behind Dyrby’s public association with the Danish #MeToo debate. It does not address Y’s AI production, copyright dispute or funding, so it supports only the biographical and controversy-related part of the supplied article.
7. https://www.ekkofilm.dk/artikler/michael-dyrby-tavs-om-sin-fortid-pa-tv2/ - In this report, Ekko examines Michael Dyrby’s reluctance to discuss his past at TV 2 during renewed scrutiny of workplace conduct and #MeToo cases. The article identifies Dyrby as a former TV 2 news director and notes his role in establishing TV 2 News in 2006. It describes questions put to him on the television programme Presselogen about whether he had been part of problems involving sexism and misuse of power at TV 2. Dyrby declined to address those questions in detail, saying the programme had invited him to discuss a separate case at B.T. The article also recounts a B.T. workplace controversy in which former student assistants alleged inappropriate behaviour by an editor; it reports that the employees were later replaced by robots, while the editor received a reprimand and remained employed. The story places these events within broader debate about leadership accountability and how media organisations investigate complaints. It states that Dyrby left TV 2 in 2015 after compromising images circulated in connection with an affair, while making clear that the contemporary discussion concerned different episodes and allegations. The report is useful context for the supplied article’s claim that Dyrby’s earlier career and #MeToo-related controversies became part of public discussion about Y. It does not establish every detail in the supplied text, including its specific reference to images being circulated to particular political representatives.