AI journalism: how news production is changing

AI in journalism (AI journalism) is the use of artificial intelligence tools to support journalists in the creation, management and distribution of news. In practice it means adopting a blended approach made of automation and human creativity: AI speeds up repetitive tasks, helps sift through vast amounts of data and suggests drafts, while judgment, source verification and editorial responsibility remain with the journalist. It is a profound shift in how news is produced and consumed, one that opens unprecedented opportunities but also demands a strong commitment to ensuring ethics and quality in information.

In this guide we look at how artificial intelligence is changing newsrooms: from automating writing to verifying sources, from personalizing the news to the risks tied to misinformation, bias and deepfakes, all the way to the role of the journalist, which remains central and irreplaceable.

What it means to bring AI into journalism

Bringing AI into journalism does not mean handing the news over to a machine, but integrating artificial intelligence into the different stages of editorial work to make them faster and more efficient. Generative AI tools such as ChatGPT, DALL-E and Stable Diffusion have expanded the scope of newsrooms, making it possible to generate article drafts, summaries, interview transcriptions and even images. This fusion of automation and human creativity offers concrete opportunities, but it also raises practical and ethical challenges around copyright, the spread of false content and public trust in the media.

The balance lies in treating AI as an assistant and not as an author. The value of journalism is not measured only by how quickly a story is produced, but by its accuracy, the context that gives it meaning and the accountability of whoever signs it. Artificial intelligence can free up valuable time, lifting professionals from the most mechanical tasks, but it is exactly that time that can be reinvested in investigations, on-the-ground verification and in-depth reporting, the work where the human contribution truly makes the difference.

How does AI change work in the newsroom?

Newsroom automation is currently the area where AI in journalism delivers the most immediate results. Newsrooms use language models to speed up the production of standardizable content and to manage workflows that, until recently, required many hours of manual work. This is not about replacing the journalist's byline, but about automating the most repetitive steps in the information pipeline, leaving the decisions that matter to people.

Among the most common applications of artificial intelligence in newsrooms are:

• Generating drafts and structured content: financial reports, sports results, weather bulletins and other data-driven formats can be produced automatically from structured sources, then reviewed by an editor.
• Transcription and summarization: AI transcribes interviews and press conferences and extracts the key points, drastically cutting processing time.
• Translation and localization: international outlets can quickly adapt content to multiple languages and markets, expanding the reach of their news.
• Optimization and distribution: headlines, summaries and social-ready versions are adapted to different channels, while data analysis helps understand which topics resonate most with the audience.
• Monitoring and research: AI tools sift through large volumes of documents, archives and news feeds to spot signals useful to an investigation.

This evolution is part of a broader transformation in the publishing sector, as we describe in our piece on AI for publishing, from information to education. In all these cases the logic is the same one that drives AI agents for business: artificial intelligence does not replace the process, it makes it faster and less burdensome, leaving editorial choices under human control.

Source verification and fact-checking with AI

If automation speeds up production, it is in source verification that AI can make a decisive contribution to the quality of information. Fact-checking has always been one of the pillars of journalism, but in the digital age the volume of content to check has grown enormously and reaction times have shrunk to a few minutes. Artificial intelligence helps journalists keep up, flagging inconsistencies, comparing different versions of a story and identifying potentially manipulated content before it is published.

AI tools can compare a claim against databases and official sources, detect images reused out of context or altered, recognize the typical signals of deepfakes and trace how a story spreads across digital channels. Organizations such as the Associated Press integrated automation into their editorial workflows long ago, experimenting with how technology can support human work without replacing it. Likewise, research carried out by academic centers such as Stanford HAI studies how to design reliable, human-centered systems.

One principle remains non-negotiable, however: the final decision to publish always rests with the journalist. AI speeds up and broadens the capacity for checking, but it does not relieve the newsroom of the responsibility to validate sources and to bear the consequences of what it publishes. The technology that enables this control is often based on targeted retrieval of information: RAG systems that embed knowledge into the models make it possible to anchor answers to documented and verifiable sources, reducing the risk of fabricated claims.

News personalization and new formats

Another frontier of AI in journalism is the personalization of the reading experience. By analyzing audience behavior and interests, outlets can offer news that is more relevant to each reader, suggest related in-depth pieces and adapt the content format to the channel and the time of day. The same story can become a detailed article, a bullet-point summary, a newsletter or a short audio piece, broadening the ways people stay informed.

This ability to adapt opens up new possibilities, but it must be handled carefully. Personalization taken to the extreme risks enclosing readers in an information bubble, showing only what confirms their views and reducing exposure to different perspectives. This is why the most careful newsrooms balance relevance with plurality, using AI to broaden rather than narrow the information horizon. The challenge is to offer the audience content closer to their interests without giving up the very function of journalism, which is also to report on what people are not searching for but need to know.

What risks does AI bring to journalism?

Fake news represents one of the most significant challenges for journalism in the AI era. If not properly managed, AI tools can accelerate the creation and dissemination of false information, making it even harder to distinguish fact from fiction. This undermines public trust in the media and requires new solutions to verify sources and ensure generated content is accurate and responsible. Combating misinformation is an absolute priority for ethical and credible AI journalism.

Beyond misinformation, there are other risks to keep carefully under control:

• Bias and distortion: models learn from the data they are trained on and can reproduce prejudices or imbalances present in that data, influencing how a topic is reported.
• Deepfakes and synthetic content: increasingly realistic artificial video, audio and images can be used to deceive the public or attribute statements to people who never made them.
• Copyright and intellectual property: using content to train models, and the generation of text and images, raise open questions about ownership and the lawful use of materials.
• Model hallucinations: generative systems can produce plausible but false statements, which become dangerous if published without verification.
• Transparency toward readers: the audience has the right to know when and how AI was used in producing a piece of content.

Addressing these risks is not only a technical matter; it concerns trust, which is the real capital of any outlet. According to analysis from the Reuters Institute, the public's relationship with the news and with new technologies remains delicate, and transparency about the use of AI is a factor that affects perceived credibility. This is why many newsrooms are defining clear guidelines on what artificial intelligence can and cannot do.

Ethics and the journalist's role in the AI era

The spread of AI in newsrooms makes the ethical question even more central. Artificial intelligence has no values, does not distinguish right from wrong and takes no responsibility: these remain exclusively human tasks. The journalist who uses AI must do so with awareness, keeping control over editorial choices, verifying what the tool produces and transparently disclosing how the technology was used. We explored these principles in our article on ethics and artificial intelligence, because the responsible use of technology is not a constraint but the condition that makes journalism credible.

Far from making the journalist redundant, AI enhances their role. The skills that make the difference, namely the ability to interview, to grasp context, to ask the right questions and to exercise critical judgment, cannot be replicated by an algorithm. Automation takes care of the mechanical work, while the higher task remains with people: to make sense of the facts, protect the public interest and safeguard the trust no machine can generate in their place. In this scenario the journalist also becomes a guarantor of the quality of AI-generated content, a role that calls for new digital skills alongside the traditional ones of the craft.

The Athics experience: the "AI in Redazione" webinar

To explore these topics further, Athics hosted the "AI in Redazione" webinar, dedicated to how artificial intelligence is transforming the journalist's role and how to address the risks and opportunities tied to these new technologies. During the event, industry experts discussed the change underway in newsrooms, from the spread of fake news to copyright protection, to the relationship of trust with readers and the prospects for more sustainable and innovative AI journalism.

Among the speakers who shared their experience:

• Andrea D’Ambrosio, Chief Editor – Sky TG24
• Letizia Olivari, Director – CMI Customer Management Insights
• Pierluigi Sandonnini, Senior Web Editor – Digital 360

The conversation between industry professionals showed that adopting AI in newsrooms is not a purely technological matter, but one that concerns culture, method and responsibility above all. It is the same approach Athics brings to artificial intelligence projects for businesses, with the Crafter.ai and Portrait platforms: putting technology at the service of people, with attention to quality, security and transparency.

Conclusion

AI in journalism is neither a threat to defend against nor a magic formula to adopt without thinking. It is a powerful tool that, used with method and ethics, can make newsrooms faster, more accurate and better able to serve their audience. Automation speeds up production, AI strengthens source verification and enables new formats, but the heart of journalism remains human: judgment, context, responsibility and trust.

If you want to understand how artificial intelligence can support your organization's processes, in the publishing world and beyond, talk to our team: we will help you identify the right use case and adopt AI in a safe and responsible way.