On July 23, the Associated Press (AP) updated its newsroom standards concerning artificial intelligence (AI). The guidelines specify that AI can be used for initial research, summarizing documents, transcribing, translating, suggesting headlines, and correcting grammar. However, it is not allowed to replace traditional reporting, sourcing, or verification processes. Each AI-generated output is thoroughly reviewed and edited by a journalist before being published. The AP strictly prohibits generative AI from creating, altering, or enhancing news photographs. Additionally, disclosure is mandated whenever AI significantly contributes to the content. The AP even extends these guidelines to AI coding assistants, emphasizing the agency’s seriousness on the matter.
The AP is not the only organization implementing AI standards. A comparative study revealed that Reuters, BBC, and The Guardian have each adopted distinct approaches. Reuters has developed tools like Lynx Insight and Fact Genie to swiftly handle financial data, requiring journalists to verify AI-generated information before dissemination. The speed of its breaking alerts is remarkable, appearing within six to eight seconds, demonstrating a balance between efficiency and accuracy.
In contrast, the BBC has proceeded more cautiously with AI integration. Its tools, such as “At a glance” and BBC Style Assist, are used to summarize lengthy articles and align regional stories with the organization’s style guide. No AI-assisted content is released without approval from a journalist. The BBC also shared research conducted with the European Broadcasting Union, revealing that AI assistants misrepresented news content 45% of the time. Consequently, the BBC introduced a label, “How we used AI,” allowing readers to discern AI involvement in the news production process transparently.
Among the three, The Guardian stands out for its evolving principles regarding AI. Initially restrictive, The Guardian updated its guidelines in March 2026 to permit limited AI uses like alt text, document analysis, and transcription under human supervision and senior editor approval. This shift signifies a cautious approach that openly acknowledges and adapts to new evidence, exemplifying a dynamic standard of practice.
However, instances of non-compliance exist, as seen in Pakistan’s Dawn newspaper in November 2025. The publication admitted to editing a business story using AI, violating its own policy that prohibited such practices. This incident underscores the importance of enforcing AI policies consistently to maintain transparency and credibility within newsrooms.
The challenge of inconsistent AI policies extends beyond individual cases. Many media outlets globally lack clear guidelines on AI integration, leading to varying levels of disclosure and accountability. The absence of standardized AI policies poses a significant risk to journalistic integrity and public trust. To address this issue, newsrooms must proactively establish and communicate their AI policies to ensure transparency, accuracy, and ethical use of AI technology.
In conclusion, transparency and accountability are fundamental principles of journalism. News organizations must openly disclose their AI policies, outlining the permissible and restricted uses of AI, emphasizing human oversight in content creation, and providing clear disclosure when AI significantly influences published material. By adopting transparent AI practices and fostering a culture of accountability, newsrooms can uphold their credibility and trustworthiness in the digital age.
