For decades, audio content in newsrooms has functioned as a “black box”—a vast reservoir of reporting, interviews and archives that remained largely invisible to search engines and difficult for journalists to navigate. For PRISA Media, the world’s leading Spanish-language media group, this invisibility represented a significant loss of potential value across its diverse portfolio of more than 50 brands, including global staples like El PAÍS and the sports daily As.
To solve this, the publisher has implemented a systemic overhaul of how it handles sound, using AI transcriptions to transform audio content from a static recording into a searchable, repurposable asset. By converting live broadcasts and archived audio into text, PRISA is effectively unlocking a massive catalog of intellectual property that was previously inaccessible to traditional text-based search tools.
The initiative is part of a broader digital transformation strategy focused on three pillars: unlocking value from existing content, increasing operational efficiency, and creating a distinct competitive advantage over generic digital platforms. This shift is particularly critical for a company that manages one of the largest audio operations globally, employing more than 1,200 journalists across radio stations broadcasting 24/7 in Spain, Chile, Colombia, and Mexico.
Breaking the ‘Black Box’ of Audio
The fundamental challenge PRISA faced was the inherent nature of audio. Because search engines are optimized for text, the insights contained within thousands of hours of radio broadcasts were essentially hidden. Diego J Pruneda Paz, the Director of Transformation, Innovation, and Voice Projects at PRISA, noted that the company sought AI solutions to improve the accessibility of both current output and historical archives.
The solution was not merely a transcription tool, but a comprehensive multimedia system. All audio is now transcribed into text and fed into a central system where journalists can instantly search for specific keywords or phrases. This allows reporters to extract precise quotes and generate summaries without the tedious process of manual listening, which can then be used to draft online articles or create social media updates.
This infrastructure was made viable through a grant from the Spanish government, designed to support the adoption of AI initiatives within media organizations. According to Pruneda Paz, the core of the project is a set of AI-powered processes that allow the company to “generate, transform, search and connect” content across various formats and different content management systems (CMS).
Operational Efficiency Across Diverse Newsrooms
The scale of PRISA’s operation presents unique logistical hurdles. The company manages a complex network of newsrooms that range from massive national hubs to tiny local offices with fewer than five employees. Previously, transcription relied on standalone solutions that required individual licenses for every user—a model that was neither scalable nor efficient for hundreds of newsrooms.
By integrating transcription directly into the company’s ecosystem, the technology has become ubiquitous. This democratization of the tool has fundamentally changed the daily workflow for journalists. For example, the system now offers transcripts as subtitles, enabling a streamlined pipeline where a journalist can combine audio, a transcript, and an image to produce a short-form video for digital platforms in a fraction of the time previously required.
Beyond Transcription: The Broader AI Roadmap
While transcription has provided an immediate win for newsroom productivity, PRISA is deploying AI across several other critical fronts to safeguard journalistic integrity and enhance user experience:
- Deepfake Detection: The company has developed an AI-powered fact-checking tool specifically for audio to identify synthetic media and deepfakes, a critical necessity in an era of sophisticated misinformation.
- Content Customization: AI is being used to refine recommendation engines, ensuring that readers and listeners receive content tailored to their specific interests.
- Technical Optimization: Beyond the newsroom, AI is being utilized for advanced analytics and coding to streamline the publisher’s backend infrastructure.
Pruneda Paz emphasized that the goal of these initiatives is differentiation. In a landscape dominated by global platforms, PRISA is leveraging AI to highlight the unique value of its professional journalism, ensuring that its high-quality audio reporting is as discoverable and flexible as its written word.
The Impact of Automated Transcription
| Feature | Traditional Workflow | AI-Integrated Workflow |
|---|---|---|
| Searchability | Manual logging / Limited | Instant keyword search across all audio |
| Repurposing | Manual transcription for articles | Automated summaries and quote extraction |
| Video Production | Manual subtitle creation | Automatic subtitles from transcriptions |
| Accessibility | Siloed licenses per user | Ubiquitous access across all newsrooms |
The transition represents a shift from treating audio as a final product to treating it as a raw material. By removing the friction associated with transcription, PRISA has effectively increased the “shelf life” of its reporting, allowing a single live interview to evolve into a radio segment, a written op-ed, a series of social media clips, and a searchable archive entry.
As the company continues to refine these tools, the next phase of implementation will likely focus on further integrating these AI-powered services across its various digital products to maximize the reach of its Spanish-language content globally. The ongoing evolution of these tools will be monitored through the publisher’s continued participation in industry forums, such as the WAN-IFRA events, where the company shares its transformation benchmarks.
We invite you to share your thoughts on the integration of AI in newsrooms in the comments below.
