By Chinedu Echianu
Most newsrooms in Nigeria are using artificial intelligence (AI) without formal policies to guide its application, raising concerns over accuracy, confidentiality, editorial accountability and the safety of journalists and their sources.
This was disclosed by the Executive Director and Chief Executive Officer of Safer-Media Initiative, Peter Iorter, at the official presentation of the Newsroom AI Policy Guidance Resources in Abuja.
Iorter said research conducted by the organisation under its Artificial Intelligence and Media (AIM) Project: Navigating the Present and Shaping the Future found a significant policy gap in Nigerian newsrooms, with most media organisations lacking formal guidelines on the use of AI.
He said the research also showed that only a small percentage of respondents had received formal AI training organised by their newsrooms, despite the widespread use of the technology.
According to him, 95 per cent of the respondents were already using AI, meaning journalists were increasingly relying on the technology without adequate institutional safeguards.
“So we’re using AI largely without anything to guide us. That is dangerous,” Iorter said.
The AIM Project, implemented by Safer-Media Initiative with funding support from the International Programme for the Development of Communication (IPDC) and coordinated by the UNESCO Regional Office for West Africa in Dakar, is designed to strengthen newsroom readiness for responsible AI adoption.
Iorter said the project had so far trained 107 journalists across 70 media organisations in Nigeria, equipping them with knowledge and practical skills to integrate AI responsibly into editorial workflows.
The three resources unveiled at the event are the Editorial Policy Guideline on Artificial Intelligence for Nigerian Newsrooms, AI Resonance Guidelines for Newsrooms in Nigeria, and Gender-Sensitive Guidance for Nigerian Newsrooms.
Veteran journalist and communications expert, Lemmy Ughegbe, who reviewed the resources, described AI as the “new oracle” of journalism, capable of delivering information at remarkable speed and with a high degree of confidence, but warned that it could also produce convincing errors.
Using an analogy of a king who relied on an oracle that correctly identified thieves nine times but wrongly accused an innocent person on another occasion, Ughegbe said journalists must always question AI-generated information before publication.
He stressed that the central lesson of the guidance materials was the need for human oversight: “Artificial intelligence may assist journalists, but it must never replace the journalist’s judgment, responsibility or conscience,” he said.
Ughegbe said AI was already being used in newsrooms for tasks including transcription, translation, generating story ideas, analysing data, suggesting headlines and, in some cases, writing opinion articles.
He, however, warned that the absence of institutional policies meant some journalists were using AI on personal devices without adequate editorial oversight.
He advised news organisations to begin with relatively low-risk applications such as transcription and translation, while exercising greater caution in investigative journalism, conflict reporting and fully automated content production.
According to him, confidential documents and sensitive source information should never be uploaded to public or third-party AI systems because doing so could expose investigations, compromise sources and create legal risks for media organisations.
He urged every newsroom to develop a clear “do not input” checklist specifying the types of information journalists must not submit to AI platforms.
Ughegbe further warned that news organisations would retain full legal and ethical responsibility for content published with AI assistance. He said AI could not be blamed when a report turned out to be inaccurate, defamatory or discriminatory because the responsibility ultimately remained with the reporter, editor and media organisation.
“All AI-generated or AI-assisted content must therefore undergo rigorous human review before publication,” he said, stressing the need for independent verification of facts, figures, quotations and framing.
The veteran journalist also highlighted the growing risks AI poses to women journalists, particularly through manipulated images, voice cloning, deepfakes, synthetic sexual content, online harassment and other forms of technology-enabled abuse.
He said news organisations should not dismiss such attacks as private social media disputes, but should provide affected journalists with institutional support, including evidence preservation, legal assistance and technical and psychological support.
The guidance also addresses the potential for AI systems to reproduce existing gender biases in journalism, including by recommending predominantly male experts for political and economic stories while associating women with health and family-related issues.
He urged news organisations to go beyond merely receiving the newly developed resources and ensure that they are implemented within their editorial structures.
“Their true value lies not in their publication alone, but in their adoption and application,” he said.
The guidance materials are therefore expected to serve as a framework for Nigerian newsrooms to integrate AI into their operations while ensuring that technology remains subordinate to truth, professional ethics, editorial independence and human dignity.
