Published on 15 October 2026 by the Knight Center for the Future of News at the Walter Cronkite School of Journalism and Mass Communication, The News + AI Field Report examines 160 AI projects developed by 101 US news organisations between 2022 and 2025.

The research, conducted by journalism scholar and media innovation expert Djordje Padejski, provides one of the first systematic overviews of how newsrooms have integrated AI into their operations. Its central finding is that news organisations have largely adopted AI to “do what they were already doing but faster”, while overlooking opportunities to develop new journalistic products, strengthen audience relationships and rethink their business models.

Automation dominates newsroom innovation

The study identifies six patterns in newsroom AI adoption. Most projects focus on automating routine editorial tasks, including summarisation, transcription, translation, drafting and workflow management. These applications improve efficiency but rarely change how journalism is produced, distributed or funded.

However, some initiatives demonstrate how AI can expand journalism’s capabilities. Automated monitoring of public meetings, large-scale document analysis and investigative research tools enable journalists to identify stories and investigate information that would otherwise be difficult to process.

Another emerging category treats “journalism as a utility”, developing AI-powered services that help audiences navigate everyday civic issues. Examples include tools providing information about property taxes, evictions, restaurant inspections and immigration. Although relatively uncommon, these experiments demonstrate how news organisations can move beyond publishing articles towards providing practical information services.

Business innovation remains underdeveloped

One of the study’s most striking findings concerns the gap between the industry’s concerns about AI and its actual investments.

Newsroom leaders interviewed for the research expressed widespread anxiety about declining website traffic, changing audience behaviour and the potential disruption of traditional revenue models. Nevertheless, relatively few AI projects focus on subscriptions, fundraising or developing alternative business models.

The research also identifies risks associated with audience-facing AI applications. While internal production tools generally operate with limited public exposure, AI-generated summaries and news chatbots can introduce factual errors and reputational damage.

Opportunities for smaller newsrooms

Despite these challenges, the report suggests that technological barriers to experimentation are decreasing. According to the researchers, “the barrier to building with AI has dropped”, allowing journalists and smaller teams to develop prototypes without extensive technical resources.

Yet AI adoption remains concentrated among major national news organisations, well-funded digital publishers and selected nonprofit newsrooms. For journalism innovators, the research highlights a fundamental distinction between adopting AI to increase productivity and using it to create new forms of public value.

The study suggests that the most promising initiatives are not necessarily those employing the most sophisticated technology, but those addressing clearly defined journalistic problems.

To make these experiments accessible, the researchers have published an interactive database featuring all 160 projects, allowing journalists, researchers and newsroom leaders to explore applications, compare approaches and identify opportunities for innovation.