07/11/2025 | News release | Distributed by Public on 07/11/2025 01:18
2025-07-11. Artificial intelligence is no longer a distant prospect for journalism. Across Latin America, newsrooms are beginning to adopt it as a practical and strategic tool - automating workflows, freeing up editorial capacity, experimenting with new formats, and strengthening their journalistic mission.
by WAN-IFRA Staff executivenews@wan-ifra.org | July 11, 2025
Over the course of three months, 16 newsrooms from eight countries in the region took part in the first edition of the LATAM Newsroom AI Catalyst - a free accelerator programme run by WAN-IFRA and sponsored by OpenAI.
The initiative aims to support media organisations in the development of functional AI prototypes that can be implemented in their day-to-day operations.
These four standout projects demonstrate how AI can be adapted to different editorial contexts and deliver meaningful, measurable results - even in small teams with limited technical resources.
In a high-stakes election year, with over 40 political parties competing in Peru, the team at El Comercio faced a significant challenge: how to investigate and monitor candidates effectively without exhausting their newsroom resources.
During the programme, they created an automated workflow using n8n and AI agents to cross-reference public datasets, identify key patterns, and uncover political connections - all without advanced programming skills.
"We managed to automate tasks that used to take weeks. With a small team, we developed a methodology for accountability journalism powered by AI," said Gisella Salmón and Mayté Ciriaco, the project leads.
Their prototype has already revealed overlapping interests and previously unnoticed political ties. The team now aims to scale the model and apply it to other areas of investigative reporting.
At Grupo OPSA, the goal was to define a strategic and ethical framework for the use of AI in editorial work. Faced with a wide array of potential use cases, the team focused on one of the most pressing: freeing up journalists and editors from repetitive tasks such as proofreading and SEO optimisation.
The result was MarIA (Autonomous Editorial Review Module), a smart assistant trained on the newsroom's editorial style guide. MarIA automatically corrects articles, suggests SEO improvements, flags missing sources, and generates structured editorial feedback - all while preserving the brand voice of OPSA's flagship titles, La Prensa and El Heraldo.
"MarIA not only streamlined our processes - it transformed our editorial culture. We moved from uncertainty to a clear, strategic roadmap for AI," the team shared.
The tool was developed by a cross-functional team, including fact-checkers, SEO analysts and data editors. By involving key editorial staff early on, they were able to reduce internal resistance and demonstrate its value. OPSA now plans to institutionalise an AI team and roll out MarIA across its newsroom operations.
El Vocero, Puerto Rico's leading free daily newspaper, was eager to explore new formats to engage wider audiences, but without placing additional pressure on its editorial staff. Their goal: automate the production of daily audio news updates with minimal human intervention.
Their solution was a fully automated workflow built using no-code platforms. OpenAI is used to generate scripts from the day's top stories, which are then sent to ElevenLabs to be converted into audio using a cloned voice of a journalist. The output is automatically mixed with branded audio and published in less than five minutes.
"We stopped treating AI as an experiment and made it part of our routine. Today, we publish a professional-sounding daily news briefing without extra effort," said Giovanny Vega, Innovation Manager.
The team plans to extend the project to themed audio segments covering sports, politics and business.
At Medcom Digital, the commercial team faced a bottleneck: producing sales presentations relied heavily on the design department, leading to delays and missed opportunities.
To tackle this, they developed ZionPath AI, a tool that automates the creation of personalised, visually compelling proposals in under 20 minutes. The system integrates smart forms with ChatGPT for adaptive copywriting, Canva's API for layout generation, and Runway for dynamic visuals.
"ZionPath cut proposal delivery times from three days to 18 minutes. More importantly, it changed our mindset - AI became a collaborator, not a threat," said project lead Kimberly McRae.
The team now aims to refine ZionPath into a scalable SaaS platform and offer it to other media outlets and agencies across the region.
Despite their varied scopes, tools and editorial aims, these four projects shared a core principle: identify a real problem and use AI to solve it in a focused, practical way.
None of the teams relied on large budgets or full-time developers. What made the difference was clarity of purpose, cross-team collaboration, and a willingness to experiment, iterate and learn.
Perhaps the most valuable lesson of all: AI doesn't replace journalism, it enhances it. By automating repetitive work, it frees up journalists to focus on what truly matters. By surfacing hidden patterns in data, it opens up new editorial possibilities.
For these newsrooms, AI is no longer a future goal, it's already shaping their present.
The LATAM Newsroom AI Catalyst is a free accelerator programme developed by WAN-IFRA in partnership with OpenAI. Its mission is to support Latin American news organisations in adopting AI to strengthen editorial workflows and strategic decision-making.
The first cohort brought together 16 newsrooms for a three-month hybrid experience that included:
Applications for the second cohort are still open. The programme is free for selected media outlets. All that's required is a committed team and a willingness to cover travel expenses for the in-person workshops.
More information and application form here:
wan-ifra.org/events/convocatoria-latam-newsroom-ai-catalyst
WAN-IFRA Staff