A new independent database tracking the use of artificial intelligence (AI) in journalism, launched in 2026, has documented over 3,700 verified and source-linked cases of AI adoption across news organisations in 144 countries over the last 15 years, providing researchers, editors, journalists, and media organisations with a growing evidence base on how AI is transforming the news industry.
Known as “J·Index” (Journalism AI Index), the initiative describes itself as the largest source-linked index of publicly disclosed cases of AI adoption in news media worldwide and maps how news organisations are using AI in editorial production, distribution, audience engagement, monetisation, and organisational operations.
Created, built, and maintained by Katya Gorchinskaya, an award-winning Ukrainian journalist, media manager, and strategy consultant, the database is developed as a beta platform that is continually updated as new cases are identified and verified. It currently covers the period 2012 to 2026.
J·Index seeks to distinguish documented AI use from promotional claims by technology companies or news organisations.
According to its methodology, each case is checked against a live source before being included in the database, and the index combines manual research, human-guided research using AI-assisted tools, selected searches conducted through other AI models, and an automated system that began in June 2026 to identify potential cases from about 20 trade-press, accelerator, and industry sources.
Potential entries are subsequently reviewed by a person and automatically checked for duplication before being added to the database.
The database records 20 structured fields for each verified case, including the organisation, country, use case, technology or tool, domain, function, level of human involvement, maturity stage, funder, and source link.
The structure is intended to allow comparisons across countries, types of media organisations, and different stages of AI adoption.
The index classifies AI applications across five broad areas of publishers’ business models, namely Editorial Production, covering the creation of journalism; Marketing, Distribution, and Discovery, covering how journalism reaches audiences; Audience and Community, concerning relationships with readers; Commercial and Monetisation, covering revenue-generating activities; and Workflow and Operations, covering the wider organisational functions behind journalism.
Within these areas, J·Index identifies more specific functions, allowing users to examine how particular technologies are being used for activities such as writing assistance, verification and fact-checking, transcription, personalisation, and other newsroom tasks.
The platform also allows users to examine the technologies newsrooms use and compare commercially acquired tools, internally developed systems, and open technologies.
One of the index’s key findings is that the use of AI in journalism is predominantly about augmentation rather than replacement.
J·Index says its field data shows that most newsrooms continue to keep humans firmly involved in their work, while only a small number have reorganised their businesses around AI.
The index distinguishes between AI systems that automate a newsroom process and those designed to automate a broader business objective.
Its current data indicates that 61 percent of documented AI cases automate a newsroom process, such as a task or series of tasks, while 27 percent are on the business side, where AI is used to pursue objectives such as increasing subscriptions, reducing audience churn, or generating more clicks.
J·Index also cites 2026 research from FT Strategies, a specialist media consultancy owned by the Financial Times newspaper, and the World Association of News Publishers (WAN-IFRA), indicating that 91 percent of newsrooms have adopted AI, while 42 percent of leaders describe the results as limited.
The index suggests that the concentration of adoption in editorial activities, rather than core business transformation, may help explain this gap.
The platform also provides a maturity framework through which users can examine how far news organisations have progressed in their adoption of AI.
At the most advanced end, J·Index identifies organisations classified at Stage 5, where AI has contributed to AI-native business models and products, as well as organisational structures designed to support them.
The Index groups some of the most advanced adopters into three broad types, namely “Transformers”, referring to established newsrooms that have substantially redistributed work between people and machines; “AI natives”, referring to organisations built without a traditional human newsroom; and “Hyperscalers”, describing AI-native organisations that use a common AI infrastructure to operate multiple local news brands.
J·Index identifies China’s AI deployment in state-run media as significantly different from patterns observed elsewhere, as its database describes a large-scale, government-driven rollout extending from national to provincial, prefecture-city and county-level news organisations.
The platform currently lists 3 national, 31 provincial, 333 prefecture-city, and 2,591 county-level state-run outlets within this framework.
The Index also highlights the extensive use of synthetic presenters and AI-generated radio operations in China and estimates that approximately one in three synthetic-presenter deployments worldwide is Chinese. It identifies a national programme under which synthetic county radio stations may be operating across all of mainland China’s 31 provinces.
The platform enables users search for AI adoption by organisation, technology, use case, or country, making it possible for newsrooms and researchers to examine how peers are approaching specific challenges.
It also identifies open-source AI projects that can potentially be reused by news organisations and documents the funding sources behind different AI adoption projects.
The J·Index database is continuously updated as new cases are identified, verified, and added, and the project says its objective is to provide an evidence base of documented AI adoption rather than relying on press releases or vendor claims. It invites users to provide feedback and submit missing entries to improve the database.
The Index can be accessed at https://jindex.ai/
