Global Media Giants Embrace AI in News Production and Distribution

AI is reshaping the news industry: from The New York Times’ data analytics and BBC’s AI translation to The Washington Post’s personalized podcasts, major international media outlets are actively embracing AI across three key dimensions—creation, product development, and distribution. While AI has significantly improved content production efficiency and product innovation, it also faces challenges related to accuracy and declining user traffic. Most media organizations emphasize that AI serves only as an auxiliary tool, and all content must undergo human review before publication; journalistic integrity and credibility remain the foundation of the industry.

Article author and source: Quanmeipai

Embrace actively while accepting challenges.

Mid-last month, Reuters appointed Jane Barrett as Head of Products, following her nearly two-year tenure as Head of AI Strategy. Internal sources commented: “Under her leadership, Reuters has actively embraced artificial intelligence—integrating AI into internal workflows and driving AI enhancements to improve user experience.”

This personnel adjustment serves as another reminder that AI is no longer an independent innovation project even in the media industry; overseas media outlets like Reuters are accelerating the integration of AI into their core operations.

When people think about AI's application in the news industry, they often imagine it assisting journalists in gathering information, proofreading content, translating articles, or even generating news stories directly. However, in reality, AI's applications have long extended beyond news production. From editorial workflows to news product innovation and distribution methods, AI is reshaping the operational model of the news industry.

Against this backdrop, this issue of MediaPai focuses on the transformation of the news industry in the AI era, drawing on practices from leading international media outlets and illustrating key industry impacts of AI across three dimensions: content creation, product development, and distribution.

We turn our attention to another media giant. Although The New York Times explicitly emphasized when introducing its AI products to readers: “We do not use AI to write news reports; all published content is the responsibility of journalists,” and stated that AI applications in news production must undergo human review, this caution has not deterred the newspaper from exploring AI.

Publicly available information shows that AI's specific applications within The New York Times newsroom include at least:

1. Number of analyses

Machine learning models process vast amounts of data to provide support. For example, in an investigative report on war conflict events, journalists used AI tools to scan satellite images to identify craters, which were then verified manually by reporters.

2. Assist in writing titles and abstracts

Editors may use AI tools to generate draft headlines and article summaries. However, all content must be reviewed manually to ensure compliance with internal AI usage guidelines.

3. Generate auxiliary translations and audio versions

Currently, most news content can be converted into audio using automatic speech technology, while Spanish-language articles are generated using AI translation models. All content is reviewed by the editorial team before publication to further expand its reach.

Regarding the use of images and videos, The New York Times emphasizes: it will not use AI to alter photos or videos, nor will it generate images intended to impersonate real news scenes. If AI-generated images are published, the publication will clearly state the reason for their use and the method of generation, and will apply an AI label to the images.

Similarly, Olle Zachrison, head of AI for BBC News, described in an interview last November how BBC internally applies AI. BBC has integrated AI into various news production processes, including large-scale transcription, content tagging, and multilingual translation, to enhance content production efficiency—AI-assisted translation has become a well-established use case. However, BBC consistently emphasizes that human review remains irreplaceable to ensure news content meets public media editorial standards.

From the practices of The New York Times and the BBC, it is evident that AI excels at summarizing and repurposing data, but a more systematic data capability framework needs to be established.

He believes that it is merely the organization and repackaging of numbers, rather than genuine news discovery.

From a news production perspective, news archives themselves are not the source of exclusive stories; the most valuable leads often come from external real-time data. Therefore, future news organizations need to enhance their systematic ability to acquire and process external data, rather than relying solely on internal news archives.

The application of AI in news media is giving rise to new product formats, with AI podcasts emerging as a key area of exploration.

In August 2025, the BBC announced the launch of My Club Daily, an AI-powered football podcast offering customized audio news for fans of five English football clubs. Specifically, the BBC used ChatGPT to generate broadcast scripts based on its own published news articles, and then employed AI to convert the scripts into audio.

BBC emphasizes that every AI-generated article and audio clip is reviewed by the editorial team to verify accuracy before publication, and users are clearly informed when content is generated using AI technology.

In the same month of December, The Washington Post also launched its AI-powered personalized news podcast, Your Personal Podcast. When using the product through The Washington Post’s mobile app, users can select their preferred news topics, hosts, and playback duration to generate customized news podcasts. Bailey Kattleman, Head of Product and Design at The Washington Post, said the goal of this AI podcast is to enable users to access news in a more flexible way.

This product was developed in collaboration between The Washington Post and the voice AI company Eleven Labs, which has previously provided audio generation services for publications such as Time and The Atlantic. Bailey, the project lead, believes this podcast format helps The Washington Post reach a broader audience, particularly younger, more diverse listeners who seek news in more engaging ways. Andrew Deck, a contributor to the Nieman Journalism Lab, also noted that this initiative represents a bold and innovative step for The Washington Post among traditional U.S. media outlets.

Specifically, the system selects and combines approximately four key news stories based on the user’s reading and listening history and topics of interest. Two AI hosts present these stories in a relaxed, conversational format, with each news item lasting no more than two minutes. The content is updated in real time to reflect the day’s latest news.

Meanwhile, as mentioned earlier, users can further customize podcast content by selecting their preferred topics, AI host characters, and podcast duration (4–8 minutes). Unlike traditional podcasts that use a one-size-fits-all approach to topics, hosts, and length, this AI-powered product generates personalized news content tailored to each user’s interests, shifting from “one-to-many” to “one-to-one.”

Image source: The Washington Post

This AI podcast product is technologically innovative and has garnered significant attention since its launch, but it has also faced criticism over content accuracy, including instances of mispronouncing journalists' names.

According to internal staff at The Washington Post, the product has exhibited various errors, including misattributed or fabricated quotes, as well as the addition of extra commentary—such as misinterpreting the source’s views as the Post’s own stance. Therefore, The Washington Post has also included a notice in the app, urging users to verify podcast information by consulting the original articles.

In addition, the union representing Washington Post employees told NPR that they have concerns about this new product and its rollout, believing it could undermine the Washington Post’s journalistic mission and the professional value of its reporters. The union also noted that the Washington Post has always corrected errors in its reporting, and therefore asked, “Why support a technology that adopts lower standards?”

Therefore, a noteworthy question is: Why are news media outlets taking the risk to launch AI podcasts?

Gabriel Soto, Director of Research at U.S. market research firm Edison Research, believes a major reason is cost advantages: AI podcasts reduce the human and material resources required in traditional podcast production, such as writers, editors, hosts, and recording studios.

Industry insider Andrew Deck speculated that if The Washington Post’s experiment succeeds, the newspaper could significantly expand its audio news offerings at a lower cost. However, Bailey Kattlema, Head of Product and Design at The Washington Post, emphasized that this AI podcast product is not intended to replace traditional podcasts. In her view, traditional podcasts still hold unique and enduring value.

From the BBC to The Washington Post, we can see that AI is driving news products from uniform, one-size-fits-all content toward personalized and interactive formats. At the same time, ensuring the accuracy of news content and maintaining media credibility have become essential challenges for media organizations in the AI era.

AI is redefining how users access news and information—it’s only a matter of time: as AI tools become increasingly integrated into people’s daily digital lives, more users are turning directly to AI platforms for news and information.

According to the 2026 Reuters Digital News Report, the number of people using AI chatbots to access news weekly is increasing, rising to 10%. Although this group remains a minority, as AI adoption accelerates, these tools are likely to become increasingly important channels for news consumption, particularly among younger audiences and those with high news engagement.

Specifically, when using AI, users not only receive news summaries but also further leverage AI to organize and analyze information, aiding their understanding of news events. This means that AI has evolved from a mere information channel into a cognitive tool that helps users gain deeper insights into news events.

This shift has precisely provided news media with new directions. For example, an increasing number of users prefer news content that is more concise and easy to understand, indicating that media outlets can better meet the needs of time-constrained audiences through clearer writing and news briefs. At the same time, users are increasingly inclined to ask follow-up questions about news topics, creating opportunities for explanatory and analytical content formats.

However, some needs cannot be easily met by news media. A major advantage of AI chatbots is their ability to quickly provide personalized responses tailored to each user’s needs—something individual news outlets struggle to achieve. Therefore, news media should not simply mimic general AI capabilities, but instead focus on their irreplaceable journalistic value.

Meanwhile, the AI platform distribution mechanism has also sparked new conflicts. In 2025, Penske Media Corporation, the U.S. media group that owns prominent outlets such as Rolling Stone and The Hollywood Reporter, sued Google, alleging that its AI summary feature used news content without authorization to generate responses, thereby impacting traffic and advertising revenue on news websites.

On the other hand, news organizations are also exploring collaboration pathways with AI platforms. For example, OpenAI has entered content licensing agreements with media outlets such as The Washington Post, the Financial Times, and The Atlantic to enable authorized use of news content on AI platforms. The Washington Post states that such partnerships enhance the reach of its news content.

Unlike changes in how news is produced and presented, as more people turn to AI platforms not only to chat and get assistance but also to stay informed about current events, what does this mean for traditional media platforms that provide news? This is likely a question that most original news providers will need to consider next.

From the perspective of news production, AI enables media outlets to produce more diverse and multilingual news content more quickly, even with limited human resources, thereby improving news production efficiency and reach. From the perspective of news products, AI also drives the development of personalized and interactive news offerings at lower costs, better meeting the needs of diverse audiences and further expanding the reach of news content.

Meanwhile, the development of AI has sparked discussions both within and outside the industry about whether AI will replace journalists. However, based on the practices of major international news organizations, for outlets that prioritize journalistic value and media credibility, AI serves as an auxiliary tool rather than the primary agent in news production. Most media organizations explicitly require that AI-generated content undergo human review before publication. Ultimately, AI’s capabilities are still grounded in human expertise accumulated over time, and its value lies in enhancing efficiency—not replacing journalists.

However, when we look at the news distribution stage, we can see that AI also presents new challenges for news media. As more users obtain information directly through AI tools, their willingness to click on links to original news articles declines, leading to reduced traffic for traditional news websites. This means news media face diminished revenue from content subscriptions and advertising, placing new pressure on their long-term business models.

Of course, challenges often come with opportunities. In response to the development of AI platforms, news organizations are exploring new collaboration models, expanding the influence of their content within the AI ecosystem through content licensing, and seeking new avenues for growth. In the future, news media must consider not only how to leverage AI to improve production efficiency and innovate news products, but also how to consistently highlight their irreplaceable professional value within the new information dissemination landscape.

Looking back at the evolution of the news industry—from the print media era, through the rise of digital portals and search engines on the internet, to the mobile social media era, and now into the AI era—each transformation in media technology has driven the news industry to continuously adapt its development model, inevitably accompanied by the pains of transition. Technology changes, production methods change, dissemination formats change, and audience habits for consuming news change as well; yet the core value of journalism remains unchanged—the accurate and timely reporting of newly occurred facts.

AI will not change the essence of news, but it is reshaping how news is produced, the forms it takes, and the environment in which it is distributed. For the news industry, what truly matters is not any specific technology or distribution channel, but rather the professional spirit, social responsibility, and credibility of journalism—these remain the foundation upon which the industry stands.