This article brings together experiences that show how different media organisations across the region are making practical decisions to integrate artificial intelligence responsibly and with tangible impact on their daily operations.
When discussing artificial intelligence in journalism, the real challenge is not the technology itself, but how to integrate it into newsrooms: which problems to address first, which tasks to automate, where to put guidelines in place, and how to do so without losing editorial focus or audience trust. Across Latin American newsrooms, this process is already underway. Far from generic solutions, media organisations are moving forward based on concrete needs, practical learning and strategic decisions that connect technology, operations and journalistic mission.
In this context, WAN-IFRA has created spaces for exchange, experimentation and applied learning, enabling media organisations to explore these technologies from within their own realities. Rather than promoting specific tools, the goal is to support newsrooms in formulating relevant questions, identifying use cases with real impact, and building shared knowledge in the face of a constantly evolving technology.
The cases presented below emerged from the LATAM Newsroom AI Catalyst, Cohort 2, a programme led by WAN-IFRA with the support of OpenAI, designed to accompany media organisations in the region in the practical exploration of artificial intelligence. Drawing from real challenges and very different contexts, participating newsrooms developed prototypes and applied solutions that show how AI can be integrated in a concrete, gradual way aligned with the needs of today’s journalism.
The real challenge is not the technology itself, but how to integrate it into newsrooms: which problems to address first, which tasks to automate, where to put guidelines in place, and how to do so without losing editorial focus or audience trust.
The Story of Tuki: Artificial Intelligence and Local Identity
Context
Diario UNO is a digital media outlet based in Mendoza, Argentina, with a sustained culture of innovation. When the team began exploring the use of artificial intelligence, they identified two clear challenges: the individual and unstructured use of AI tools within the newsroom, and the heavy workload of low-value tasks such as transcription or rewriting, often taken on by highly experienced journalists. From that need, Tuki emerged, with the aim of bringing structure to AI use and freeing up time for more in-depth journalism.
The Project
The development of Tuki began with a prototype focused on converting audio from Radio Nihuil into draft news articles. Over time, it evolved into a tool accessible to journalists across the group.
This growth also brought an organisational challenge: moving from experimentation to real newsroom use required coordination between editorial and technical teams, and understanding implementation as a cultural change, not only a technological one.
Results and Learnings
Today, Tuki enables the generation of draft articles from audio and written documents, incorporating the outlet’s style guide and internal standards. From the start, the team maintained a clear human in the loop approach: automation acts as a layer of efficiency, while journalistic judgement and human editing remain central.
The main learning was systematisation. AI stopped being a dispersed practice and became a shared process, with clear rules and objectives.
What Comes Next
The next step is to scale the experience and consolidate Tuki as a broader editorial support platform, reinforcing a more efficient and structured way of working.
From Intuition to Editorial Intelligence: Integrating AI into the Editorial Routine
Context
At Grupo La Silla Rota, the conversation about artificial intelligence did not start from technological enthusiasm, but from a concrete discomfort. As an independent multimedia group, with outlets such as La Silla Rota, its regional editions, SuMédico and La Cadera de Eva, the team produced large volumes of editorial content. However, many key decisions still relied on intuition and scattered signals. The data existed, but arrived too late to be part of the daily conversation.
The Challenge
From the outset, the team was clear that AI would only be useful if it could be integrated into the editorial routine without friction. With that goal, AURA was created — a prototype designed to be present before planning meetings, bringing context, signals and relevant trends without altering the team’s everyday dynamic. AI was not meant to replace decisions, but to inform them better and in time.
AI was not meant to replace decisions, but to inform them better and in time.
The Process
AURA’s development was collaborative and incremental. Editors, analytics and technical support worked in short cycles, prioritising real usefulness over technical sophistication.
Results and Learnings
The initial results helped turn isolated metrics into a shared conversation. Data stopped being late reports and became a common starting point for discussing topics and editorial priorities. The biggest learning was a shift in perspective: AI moved from being seen as a distant promise to being understood as infrastructure that supports journalistic judgement.
What Comes Next
The next step is to continue refining AURA and strengthen the editorial team’s analytical capabilities, integrating artificial intelligence ever more naturally into decision-making.
LIZA: The AI Assistant that Restores Memory and Time in the Newsroom
Context
In the wake of the rise of artificial intelligence, Primicias began exploring how this technology could solve a concrete newsroom problem: the time journalists spent searching for historical information to provide context for current reporting. Two factors intensified this challenge: the absence of a consolidated SEO strategy and an inefficient internal search tool.
The goal was clear: to reclaim time for reporting, finding different stories and deepening coverage.
The Project
During the process, the team defined the challenge around knowledge management and time. From there, LIZA emerged — a generative AI prototype conceived as a search and writing assistant. The tool was trained on the outlet’s historical content, primarily from the Politics and Economy sections, and on its own editorial guidelines. In addition, external sources were integrated to detect relevant facts in real time, thanks to the joint work of an interdisciplinary Editorial and Technology team.
Results and Learnings
LIZA consolidated into a tool with three main functions aimed at reducing production time: contextualisation based on the outlet’s archive, draft generation under editorial standards, and SEO support through headline and linking recommendations. The key learning was confirming that the technology must adapt to the editorial problem — not the other way around. Building a bespoke solution also helped preserve the outlet’s identity.
What Comes Next
The next step is to expand LIZA’s use across all newsroom sections and form a group of journalists to test the tool more intensively. In the medium term, the team is exploring extending its use to audiences and strengthening the project through partnerships and specialised AI funding.
Taken together, these cases show that the adoption of artificial intelligence in Latin American newsrooms does not follow universal solutions, but rather informed decisions grounded in practice. In different contexts, media organisations started from real problems and progressed through gradual approaches, prioritising usefulness over technical complexity. AI appears not as a replacement for journalistic judgement, but as supporting infrastructure that brings order to processes, frees up time and expands capabilities. More than the technology itself, what makes the difference is clarity of purpose and a willingness to learn along the way. At the intersection of experimentation, editorial culture and collective learning, new ways of practising journalism in the region are beginning to take shape.
About WAN-IFRA and Upcoming Spaces for Exchange
In line with these experiences, WAN-IFRA will continue offering conferences and regional programmes where media organisations can go deeper into topics such as applied artificial intelligence, editorial innovation and the sustainability of the journalism business. More information about these spaces is always available through our regional channels and on the events page.