Newsrooms Embrace AI for Investigative Journalism

Approximately 150 journalists, researchers, and technologists convened at The New York Times Building late last month for a forum focused on the evolving role of artificial intelligence in investigative journalism. This gathering highlighted the growing adoption of AI tools within newsrooms, a trend underscored by the fact that a record eight Pulitzer Prize awardees this year disclosed their use of AI to the judging panel. These journalists leveraged large language models (LLMs) for critical tasks such as organizing, summarizing, analyzing, and translating extensive document collections, significantly enhancing their investigative capabilities. The New York Times, a co-host of the event alongside the nonprofit organization Hacks/Hackers, has been at the forefront of this integration. Juliana Castro Varón, the Times’ senior design editor of AI initiatives, emphasized that the newsroom's approach is not about "robots telling the news" but rather about employing AI to "make the work we do more efficient, more effective, and able to do things that otherwise would be very, very hard to do." She further noted that AI's strength often lies in its capacity to understand images and text, rather than solely generate them. The Times’ AI Initiatives team, comprising nine journalists and technologists, spearheads the newsroom's AI strategy and deployment. Hacks/Hackers, known for fostering collaboration between journalists and technologists, has previously organized similar summits on AI and journalism in Baltimore over the past two years. The forum featured a series of workshops, panels, and presentations exploring the potential benefits of AI for investigative reporting, delivered by prominent advocates within the news industry. Despite the enthusiasm, discussions also addressed the inherent limitations of current AI tools, including the persistent issue of "hallucination" in LLMs, where the models generate factually incorrect information. The Pulitzer Prize for Investigative Reporting awarded this year recognized a New York Times team for their audit of the Securities and Exchange Commission's (SEC) cryptocurrency lawsuits, which revealed a pattern of weakening enforcement during the second Trump administration. This award-winning reporting was notably assisted by LLMs. The forum aimed to provide a platform for sharing best practices and addressing challenges as AI becomes an indispensable tool in the pursuit of in-depth journalistic investigations, enabling reporters to tackle complex data sets and uncover stories that might otherwise remain hidden.