AI enters Bosnia's newsrooms faster than editorial rules

Artificial intelligence is already part of the daily work of journalists in Bosnia and Herzegovina. It is used for interview transcription, translation, searches, document summarization, photo editing, headline preparation, and audience reaction analysis.

However, some newsrooms still do not have written rules clearly defining which tools may be used, what information must not be entered into them, who must be informed when AI is used, and when audiences should be notified, CE Report quotes FENA.

Responses from media outlets, regulatory and self-regulatory bodies, a fact-checking platform, and experts show that the technology has entered newsrooms in BiH faster than editorial and security procedures have been developed.

In practice, decisions on how AI is used are therefore often left to individual journalists, who rely on their own judgment, professional standards, and the terms of service that most users probably never read in full.

This raises questions that are no longer merely technological: Who is responsible when AI fabricates a fact or quote? Can an unpublished interview be entered into a chatbot for transcription? Must an AI-generated illustration be labeled? What happens to confidential source data? Can an editor monitor the use of tools they were not informed about?

Rubina Čengić, a journalist and editor at Objavi.ba, says their small newsroom has no internal guidelines and instead relies on recommendations from the European Federation of Journalists.

She uses ChatGPT for research, always requesting relevant sources and links, as well as for translation, correcting typographical errors, consulting on headlines, and creating illustrations when suitable photographs are unavailable. She uses TurboScribe for transcription.

However, she draws a clear line between technical assistance and journalistic work.

“If we are talking strictly about journalism, I would never use artificial intelligence on its own for fact-checking, analysis, summarization, or writing. If I ask for a headline suggestion, that is more of a consultation when none of my colleagues are nearby,” Čengić said.

She uses her own free user account for this work. She believes journalists should always inform their editor when they have used AI and that any content in whose creation the tool participated should be labeled.

In her view, the label can be a simple sentence stating that AI was used for research, illustration, or another part of the work.

“It is important to know who is behind a journalistic product. This is important for trust between the media and its audience, and trust needs to be built, maintained, and protected,” she stressed.

A similar situation exists at the public broadcaster BHRT, where no written guidelines, regulations, or other internal procedures governing the use of AI tools have yet been adopted.

“As a result, neither the tools employees are allowed to use nor those that would be prohibited have been defined. Colleagues in Multimedia, and probably in other newsrooms as well, occasionally use AI for transcription, translation, research, document summarization, photo editing, and headline preparation, depending on their own needs and preferences,” BHRT said.

Due to its prolonged financial crisis, BHRT does not have paid business versions of AI services. Journalists use free tools, most often different versions of ChatGPT, through their own accounts and at their own responsibility.

One editor in the News Programme pointed to the broader context, noting that even the computers journalists work on are not adequate for their jobs, let alone that the organization could consider paying for AI tools.

“Editors expect to be informed when a journalist uses artificial intelligence, but this obligation has not been formalized. BHRT currently does not produce content that is significantly generated or altered by AI, nor does it use synthetic voices, virtual presenters, or AI-generated photographs and videos,” the public broadcaster said.

Although it has no official guidelines, BHRT says it discourages entering unpublished interviews, audio recordings of conversations, court files, personal data of interviewees, and confidential-source information into publicly available tools.

Responsibility for published content remains with programme editors and sector directors, in accordance with existing internal rules and employment contracts. The professional standards they follow require all information to be verified, including information suggested or generated by AI.

Raskrinkavanje.ba, a platform that fact-checks information published in the media space, also assesses that AI applications and tools are already widely present in domestic newsrooms, although without detailed research it is impossible to reliably determine the extent of their use or the tasks for which they are used.

AI can reduce the time needed to perform certain tasks, the platform stresses, but it cannot replace information verification and fact-checking. Rules governing its use remain, for now, a matter of editorial policy that each newsroom establishes independently.

One of the most serious problems arises when a journalist enters an interview, recording, court file, contact details of an interviewee, or a document obtained from a confidential source into a publicly available chatbot.

Bosnia and Herzegovina's Personal Data Protection Agency says that simply entering such content into services such as ChatGPT, Claude, Gemini, or Copilot constitutes the processing of personal data and, in most cases, its disclosure or transfer to a third party.

Under Bosnia and Herzegovina's Personal Data Protection Law, processing encompasses almost any operation performed on data, from collection, recording, and storage to structuring, use, transfer, disclosure, combination, and deletion.

This means that a journalist is processing data from the moment they “paste” a file or text into a chatbot window, because they collect and structure it for input, transfer it to an AI system, and use it to obtain a transcript, summary, or analysis.

The law does not specifically regulate the use of artificial intelligence, but it applies to automated and non-automated processing of personal data. As data controllers, media organizations must therefore take technical and organizational measures to protect data and assess the risks that AI tools create for the people whose data is being processed.

The Agency particularly warns that even obtaining an interviewee's consent does not automatically resolve the issue.

“When there is no other legal basis, consent can be used for processing. However, even after obtaining consent, the media outlet remains responsible for data security. If it enters data into a free or commercial general-purpose chatbot that has no data-processing agreement and uses user inputs to train the model, the media outlet may make an unauthorized disclosure of data to a third party, regardless of the fact that the interviewee agreed to the use of artificial intelligence,” the Agency said.

The Agency states that journalists must not enter unpublished court, police, medical, or other confidential documents containing personal or sensitive data into free AI tools for summarization, translation, or analysis.

Even when there is a legal basis for processing such data, it may only be carried out through systems that guarantee appropriate technical and security measures and protect submitted materials from leaks and unauthorized access.

Raskrinkavanje takes a similar position.

It considers entering unpublished interviews, source information, court files, and confidential documents into publicly available AI tools extremely risky because such services may store submitted data, use it for future training, or expose it to other people as a result of a security breach or hacking attack.

Zarfa Hrnjić Kuduzović, an associate professor at the Faculty of Philosophy at the University of Tuzla, stresses that AI tools can compromise source security, particularly when unpublished interviews and documents are entered into commercial or free services.

“The content of a prompt is stored on the servers of foreign IT companies. If a journalist enters a sensitive document or unpublished interview into a free chatbot, they risk a data leak and directly violate their journalistic obligation to protect the identity of their sources,” she warned.

She recommends that journalists disable data-sharing for training in the tools' settings, while emphasizing that this is no substitute for caution and risk assessment.

Those interviewed were almost unanimous that responsibility for inaccurate or manipulative information cannot be shifted onto an algorithm.

The Press Council and Online Media Council of Bosnia and Herzegovina states that the publisher, editor, and author bear professional and legal responsibility for AI-generated content.

AI tools have no legal personality and constitute a technical means. The fact that a chatbot produced an error cannot reduce the responsibility of the media organization that decided to publish the content.

The Council notes that the Code already establishes editorial responsibility for texts taken from other media, even when the source is clearly identified. The editorial decision to republish something carries its own responsibility, and the same rule must apply when content originates from an AI system.

Rubina Čengić also stresses that journalists and newsrooms are responsible for everything they publish.

“A journalist's job is to verify information, including information obtained from artificial intelligence. The emergence of artificial intelligence has not abolished professional codes,” she said.

Hrnjić says artificial intelligence is not a conscious actor but a software tool, meaning legal and professional responsibility remains with humans.

In her view, publishing unverified information generated by AI constitutes professional negligence, even when the journalist had no intention of misleading the public.

The Communications Regulatory Agency states that professional, editorial, and legal responsibility lies with whoever publishes the information. Every broadcaster is responsible for all content in its programme, regardless of the source of the information or the technology used to produce it.

Raskrinkavanje says responsibility for an inaccurate fact obtained with the help of AI is the same as responsibility for any other inaccurate information that is published.

Those interviewed agree that content that realistically depicts something that did not happen must be labeled. There are differences, however, over whether audiences should be informed about every use of AI, including technical tasks such as transcription and spelling corrections.

In their view, the use of AI for translation or technical photo processing, followed by verification by the author and editor, does not necessarily need to be disclosed. However, AI-generated illustrations and photographs, as well as significant portions of AI-generated text, must be clearly labeled.

“When audiences discover that media content was synthetic and this was not disclosed, distrust spreads even to completely credible and authentic journalistic stories. This undermines the credibility of journalism as a whole,” Professor Hrnjić warned.

The Communications Regulatory Agency likewise considers AI-generated photographs, voices, videos, and virtual presenters without clear labeling potentially misleading.

The existing Code for Print and Online Media in Bosnia and Herzegovina does not sufficiently regulate the use of artificial intelligence, but the Press Council and Online Media Council is working on amendments.

Two meetings of an expert group were held in April and June 2026. Editors, journalists, lawyers, and university professors analyzed the ethical challenges posed by AI and prepared a draft of the amendments.

A new article of the Code is expected to regulate the ethical use of AI in print and online media, with the amendments expected to be finalized and presented to the public by the end of the year.

Hrnjić says that in news journalism, AI speeds up routine tasks, summarizes agency reports and documents, adapts text for different platforms, organizes notes, and helps develop story ideas.

In investigative journalism, it can assist in analyzing large databases, searching thousands of pages of documents, filtering information, mapping networks of connected individuals, and analyzing satellite imagery.

However, the greatest risks are convincing hallucinations and algorithmic biases embedded in models.

“An output produced by artificial intelligence should be treated solely as a lead that must be verified,” she stressed.

Another risk is the weakening of critical thinking.

If journalists routinely allow AI to connect causes and consequences for them, their ability to conduct in-depth analysis and question information could weaken. Hrnjić notes, however, that this is a relatively new technology and that long-term empirical evidence is needed before drawing definitive conclusions.

Rubina Čengić expects AI to affect the number of employees and the distribution of work to some extent because it makes news production cheaper. Nevertheless, she believes the essential work of journalists cannot be automated.

“The main job of a journalist is verification, and no person or thing can replace journalists in that,” she said.

BHRT does not currently expect an immediate impact on employment. According to one of its assessments, more intensive use of AI tools could primarily increase the speed and productivity of work.

Those interviewed agree that journalists and editors will need training in AI literacy, data protection, fact-checking, and copyright.

While newsrooms are already using artificial intelligence, common rules are still being prepared and internal policies often do not exist. Until they are adopted, decisions about what is permitted, what is risky, and what audiences must know will largely remain with journalists and editors.

Photo AI