Academics: AI Presenters Not Yet Effective for Television News
A communications lecturer at Budi Luhur University, Muhammad Ikhwan, stated that the use of artificial intelligence (AI)-based presenters in the television news industry still faces a number of obstacles and cannot yet replace production needs that demand high speed and accuracy.
During a discussion held at the Antara Education Institute in Jakarta on Thursday (11/6), he said his research on the implementation of AI presenters at a private television station showed the technology is not yet effective for hard news and breaking news coverage.
“Television requires speed. When an event must be broadcast immediately, the production process for an AI presenter still takes quite a long time, making it ineffective for breaking news needs,” Ikhwan said in Jakarta on Thursday.
Additionally, the AI development team’s separation from the newsroom makes production coordination more complex. “To produce content that meets programme requirements, coordination between the editorial team and the development team is necessary. This process takes time and is ultimately considered inefficient,” he stated.
Besides technical factors, limited human resources and high operational costs are also challenges in developing the technology. He mentioned that managing an AI presenter requires various paid software, most of which use subscription models in US dollars, thus adding to the cost burden for media companies amidst a television business climate already under pressure.
Another finding in the research showed that the audience response to AI presenters tends to be less positive. This is evident from the rating performance of programmes featuring virtual presenters. “When AI content appears in a news programme, the ratings tend to drop. Viewers still seem to expect news to be delivered by a human because it is considered closer to facts and reality,” Ikhwan said.
In addition to researching AI presenters, Ikhwan also examined the use of AI to visually reconstruct criminal cases in television programmes. He assessed that AI technology still has limitations in accurately depicting facts because the resulting visuals do not always correspond to the actual conditions. “No matter how good the command or prompt given, an AI illustration cannot fully represent reality. That is where the issue of accuracy arises, which is an important principle in journalism,” he said.
According to Ikhwan, using AI for criminal case reconstruction also raises ethical challenges related to privacy, copyright, protecting the identity of victims and suspects, as well as the potential for public judgement through the media. He assessed that the development of AI technology in broadcast media must be followed by more specific guidelines so that technological innovation remains aligned with journalistic principles. “AI does offer many opportunities, but its use must still maintain accuracy, balance, and journalistic responsibility,” Ikhwan said.