This past year has been professionally gratifying, as larger local news organizations have begun to take AI adoption seriously. They have refocused internal resources to move beyond mere experimentation, integrating AI-powered products into meaningful workflows. These well-resourced newsrooms now approach AI adoption with the discipline derived from formal product management principles and the software development lifecycle.
Looking ahead, I predict that 2025 will mark a turning point for smaller local newsrooms with limited resources, as they begin to formalize their approaches to AI product adoption.
Why is formalizing AI adoption necessary? Experimentation, while valuable, has its limits. To make a substantive organizational impact, AI must be widely deployed and integrated into journalists’ daily workflows. Achieving this requires addressing several critical questions: How will AI products be supported and maintained? What processes will guide the identification and adoption of new features? And how will the operating expenses be sustainably managed?
Fortunately, much of this formalization need not be novel. Most newsrooms already have established processes for technology adoption that predate AI. These frameworks align stakeholders on a problem, determine viable solutions, and implement them effectively.
However, generative AI disrupts standard adoption processes by making immensely powerful technologies directly accessible to journalists at large. And they have certainly been experimenting with generative AI, as we saw in our research published this year. While these efforts are both commendable and necessary, they also reveal a challenge: most journalists lack formal training in product management or the software development lifecycle. News organizations must now work to translate promising experiments into formal products that can be shared responsibly with others.
Thankfully, many newsrooms already have the infrastructure to meet this challenge. Local journalists conducting groundbreaking AI experiments should collaborate with colleagues experienced in product development and the software development lifecycle. Together, they can leverage existing processes to formalize the adoption of new AI products, maximizing their chances of success and ensuring they are implemented responsibly within news operations.
Ernest Kung is the AI product manager at the Associated Press.
