Ioana Manolescu: Using AI to rebuild trust in journalism
Every day, high-quality public data is collected around the world, from health and education to agriculture, economics, and the environment. But for journalists and citizens alike, finding clear answers in these vast, fragmented datasets can feel impossible. With the PREDIAL project, Hi! PARIS chair Ioana Manolescu is building AI tools that make data exploration faster, more accessible, and more trustworthy.
“Much of this information already exists, hidden in statistical portals and databases,” she explains. “Our goal is to help journalists access it quickly and trace it back to the original source.”
From complexity to clarity
Whether you’re looking for EU subsidies for crops across countries or student performance by region, official data exists, but often lives scattered across platforms, buried in dense tables or technical formats. PREDIAL extends tools like StatCheck, an AI-powered assistant developed by Manolescu’s team, that can surface precise answers in seconds.
“Instead of spending hours clicking through statistical websites,” she says, “a journalist can now ask a question in natural language and get reliable, sourced information instantly.”
It’s not just about speed. It’s about transparency. Every answer given by PREDIAL is linked to its original source, allowing readers and reporters alike to verify the facts. In a time when disinformation
Rethinking the reporter’s toolbox
AI is already transforming many fields, and journalism is no exception. For investigative reporters, AI tools can connect dots across hundreds of documents, people, organizations, and timelines that no human could manually piece together.
PREDIAL isn’t just a search engine. It’s designed to support discovery: suggesting questions a journalist might not have thought to ask, surfacing new leads, or highlighting anomalies that deserve a second look.
“We want to build tools that don’t just find answers, but also spark curiosity,” says Manolescu.
Training future journalists
Just as historians once spent months sifting through archives, journalists today are navigating a digital flood of structured and unstructured data. In the next decade, Manolescu believes that tools like those developed in PREDIAL will become standard in every newsroom.
“We’re not replacing the reporter’s judgment or ethics,” she emphasizes. “We’re augmenting it with faster, deeper access to high-quality sources.“
Future journalists trained with tools like PREDIAL will not only retrieve facts more efficiently but also discover new angles and connections. The result? Smarter stories, better informed citizens, and a media ecosystem better equipped to handle complexity.
Building trust with every query
In an age when AI can generate convincing but fake content, provenance, the ability to trace where information comes from, has never been more critical. PREDIAL tackles this directly: every result is anchored in transparent, publicly available sources.
“We need to reestablish a common ground for facts,” Manolescu says. “When journalists use tools like PREDIAL, their audiences can see where every claim comes from. That’s essential for rebuilding trust.“
Her work has already found a real-world testing ground: collaborations with journalists from Radio France, where PREDIAL’s tools are being deployed to support data-based reporting.
Data, democracy, and the public good
At its core, PREDIAL is about empowering informed decision-making. Whether it’s for a policy debate, a voting decision, or a public investigation, access to reliable data strengthens democracy.
“The data is there,” says Manolescu. “We’re just making it easier to reach, explore, and understand, so that everyone, from journalists to citizens, can use it to ask better questions and find better answers.“
More profiles of our 2025 chairs will follow soon, exploring how AI is reshaping research, media, and the way we understand the world.