The Chinese state news agency Xinhua has started an international project to set the rules for the proper application of AI in journalism, which has gathered 22 media bodies and professional institutions from various countries. In a world where more newsrooms are embracing AI technology, the plan seeks to tackle escalating concerns about misinformation, AI-generated content, data security, and the job of human news reporters.
The initiative was announced on November 7 at the seventh World Media Summit in Beijing, and will take place over a two-year period to work out practical guidelines for the use of AI in the global media industry. Xinhua does not wish various institutions to develop their own individual strategies, but to do so on the basis of a common set of rules, in the application of AI.
Xinhua Wants Media Organizations to Agree on Common AI Rules
The project is called the Global Media Initiative for the Joint Development of AI Application Standards, and involves 22 organizations, including 11 Chinese and 11 international ones. Subscribers are top-tier media outlets like People’s Daily (China), China Media Group, TASS (Russia), Bernama (Malaysia), Qatar News Agency, Independent Media (South Africa), universities, and specialized institutions.
The program will be put into practice in two years – including research, practical testing, and international cooperation – Xinhua reports. It will also develop a global repository of AI applications across the media to learn from each other and share examples. The goal is to develop guidelines on Artificial Intelligence that can be implemented in different newsrooms, considering different countries’ needs.
Not just simple AI News labeling
Its goals are to foster transparency and accountability in the use of AI in news production. It recommends to indicate AI generated/assisted content, to verify information with reliable sources, to test AI models for accuracy, to protect confidential data and manage risks such as deepfakes, bias and misleading outputs etc. It also suggests the use of documentation that includes the prompts, source materials, editing changes, and verification procedures that are used in the creation of AI-generated content.
This document is used to emphasise the importance of ensuring that AI-generated content can be used for drafts or reference content, but not automatically published. There should be no reason for journalists and editors to cease to have responsibilities for fact-checking and approving news. In addition, it will ban any quotations generated by artificial intelligence, and adding fake events or fake images and video created from these elements in quotations should be labelled as such and not presented as actual footage.
Why AI Standards Are Becoming More Important for Newsrooms
AI is being utilized in various creative ways across today’s newsrooms, from summarising documents to analysing information, such as transcribing interviews, translating reports, drafting initial articles, and assisting journalists in analysing data. They can be useful to automate repetitive tasks and boost efficiency, but they can also contribute to misinformation, realistic deepfakes and AI-generated content. Others may face challenges, such as those for media outlets which may not have a dedicated fact-checker or advanced fact-checking tools.
Shared standards would help improve the newsrooms’ capacity to meet these challenges and could foster standards for the verification of information, international knowledge exchange, and provision of training opportunities. The programme’s concept is based on a single goal: to make AI tools a useful ally—and never a replacement—for journalists’ decision-making and editorial choices.
Google’s AI Classroom Policy Leaves Schools Facing a Difficult Choice
The Biggest Question Is Whether These Standards Will Be Widely Adopted
The scheme has been introduced by Xinhua, but there are some questions regarding how it will be implemented. The framework is not compulsory, and those organisations that join can adapt the recommendations of the framework to their needs. How the standards will be updated, evaluated, and embraced by newsrooms outside of the early adopters, however, remains to be seen. But how they will impact others will depend on practical guidelines, an easy testing process, and resources that smaller media companies can access.
The launch was held at the 7th World Media Summit in Beijing (5-9 November) with the theme of Media in the Intelligent Age: From Consensus to Action. The event was attended by some 500 participants from 330 media outlets, think tanks, and governments in 112 countries and regions, Xinhua reported. Successful implementation of the initiative will be determined by the ability to implement the recommendations as standards that build on accuracy, transparency, and public trust, and enable the support of journalism through AI.
