Norwegian news agency NTB published an article about Telenor's security report containing false quotes and a fictional security director, which they later retracted after admitting improper use of AI tools in their reporting process.
On October 15, Norwegian news agency NTB published a news report about Telenor's annual security report that contained significant factual errors. The article included five direct quotes that could not be found in the original Telenor report, and incorrectly attributed statements to a fictional security director named 'Håkon Berg' who does not exist at Telenor. The actual report's foreword was written by CEO Birgitte Engebretsen, not the fictional director. NTB's news editor Christina Dorthellinger Nygaard confirmed that AI was used improperly in creating the article, violating journalism standards. The erroneous article was quickly retracted the same morning and replaced with corrected content. However, some news outlets like Finansavisen continued to display the false information. NTB conducted an internal review and found no other similar incidents. The agency has internal AI tools including a 'quote helper' but confirmed that external AI tools were used inappropriately in this case. Telenor contacted NTB to point out the errors and NTB apologized for the mistakes.
Domain classification, causal taxonomy, severity scores, and national security assessments were LLM-classified and may contain errors.
AI systems that inadvertently generate or spread incorrect or deceptive information, which can lead to inaccurate beliefs in users and undermine their autonomy. Humans that make decisions based on false beliefs can experience physical, emotional or material harms
AI system
Due to a decision or action made by an AI system
Unintentional
Due to an unexpected outcome from pursuing a goal
Post-deployment
Occurring after the AI model has been trained and deployed