The pervasiveness of algorithmic systems, particularly boosted by advancements in Artificial Intelligence (AI) and it Generative models, is reshaping journalistic practices and news value chain. However, journalists’ ability to critically engage with these technologies remains underexplored. Building on multidimensional models of algorithmic literacy (Dogruel et al., 2021, Swart, 2021) and AI literacy frameworks (Deuze & Beckett, 2022; Ng et al., 2021) this study investigates how Portuguese journalists perceive and interact with algorithmic systems, addressing a critical gap in current research. Employing a mixed-methods design, the study integrates an online survey of 219 journalists with three focus groups to capture both quantitative and qualitative aspects of the cognitive, affective and behavioural dimensions of the Portuguese journalist’s algorithmic literacy. The results reveal a complex and multifaceted panorama. While many journalists demonstrate an overall awareness of algorithms, they lack a deeper knowledge, with confusions between AI and algorithms operations. The affective dimension highlights a tension between the acceptance of digital tools as indispensable to remain competitive and concerns with issues like filter bubbles in social media and transparency in GenAI tools. Behaviorally, we identify that most journalists declare moderate skills in using social media, and lower skills in the use of Gen-AI tools. The results also identify a lack of structured training and newsroom support for adopting new tools, particularly Gen-AI. Based on the results we propose a framework for targeted professional development, outlining four key areas for intervention: (1) Foundational Knowledge, (2) Algorithmic Awareness, (3) Technical Proficiency, and (4) Ethical Standards. The study concludes that strengthening algorithmic literacy is vital for enabling journalists to navigate and critically engage with increasingly algorithm-driven media ecosystems.

--