She is bright, gorgeous, ageless, tireless, speaks multiple languages and is completely under my control, said the India Today Group’s vice chairperson, Kalli Purie, at the April launch of “Sana,” the AI-powered news anchor with a human-like appearance who is fed data that it can read using text-tospeech technology. News anchors are the new avatars of AI. Sana is not unique, however. At the time this article was being written, there were an increasing number of AI-generated news anchors gracing screens across the country, including “Lisa” of the private news channel Odisha TV,
“Soundarya” of the Kannada news channel Power TV, and “AI Kaur” of News18 Punjab/Haryana. According to “robot anchor” Soundarya, “AI is leaving its footprint in every industry, and the TV news industry is no different.” But are these AI anchors revolutionising TV news and journalism? Whatever the case, this was bound to happen in AI’s Hogwarts, given the recent wave of generative AI. Since the launch of the chatbot ChatGPT in November last year, AI-generated news presenters have been steadily gaining popularity. Will they be able to fully replace human anchors, though? “Zhang Zhao,” an English-speaking “male” news anchor powered by AI, was first introduced by Xinhua, China’s state news agency, in 2018.
As part of the 2022 Winter Olympics in Beijing, China also debuted a virtual host and the first AI presenter of sign language in the country. China’s stateowned news outlet People’s Daily launched their new AI-powered “female” presenter this year under the name “Ren Xiaorong,” stating that Xiaorong possesses the professional abilities of a “thousand presenters.” As Xiaorong put it, “Every conversation, every feedback you provide, will only make me smarter.” Newer avatars would also become smarter and more efficient thanks to technology. And similar bots are now in action in many other countries. “Snezhana Tumanova” was announced as Russia’s Svoye TV’s first virtual weather presenter in March 2023. An AI news presenter named “Fedha” was unveiled by Kuwait News in April 2023. A Taiwanese news broadcaster unveiled an AI weather anchor on YouTube in July 2023. And so on. Initially, it’s sort of magic that is induced in the spectators through an AI anchor. The channel authorities could wish to take advantage of that. However, that advantage would eventually be lost due to the proliferation of AI news anchors on various channels.
But the issue remains: Is an AI news anchor a milestone or a menace? And there are important dimensions for society that are not immediately apparent. Some experts worry that at least in some countries AI presenters may eventually be used to spread propaganda. The AI news anchor for China’s state-run publication, “Ren Xiaorong,” solely addressed preset questions about preset topics in relation to the “Two Sessions” conference in March. These include “cutting-edge technology,” “tax and fee reduction,” “employment,” and “The Belt and Road Initiative.”
Her responses, understandably, promoted the Central Committee of the Chinese Communist Party. One could contend that an AI presenter is simply doing what the majority of genuine presenters would do, which is adhering to directives from employers. In the beginning of 2019, the Russian official news channel Rossiya 24 presented a robot presenter named “Alex,” who caused controversy after some viewers claimed that he was spreading political propaganda. The Venezuelan state-owned television station VTV was accused of using deepfake English-speaking hosts from a fictitious American news agency to share falsely positive news coverage about the country. It is obvious that AI has the capacity to purposefully spread false information.
Additionally, it is noteworthy that a February 2023 New York Times article by foreign tech correspondents Adam Satariano and Paul Mozur regarding online misinformation was titled “The People Onscreen Are Fake. The Disinformation Is Real.” Thus, it may be argued that, at the very least, a clear regulatory framework for AI news anchors needs to be established. This March saw the debut of NewsGPT, the first news channel in the world whose content is generated entirely by AI. For some media workers, it must have been a threat to their jobs. Is AI now eating up news anchors’ jobs as well? As messages are typed into the system, AI anchors will report “tirelessly” 24×7, from anywhere, and may read out news without any interruptions. In 2018, when Zhang Zhao with AI support was first unveiled, Xinhua asserted that it would cut the “cost of news production and improve news efficiency.”
Therefore, one may infer that introducing AI news anchors to the channels would be cost-effective. Of course, AI anchors can function continuously until they are replaced by an updated version. Additionally, there won’t be any disagreement between an AI anchor and the channel’s management regarding any special or customized news they may be requested to broadcast. There is growing concern about their impact, though. And one thing is certain; journalism is also going through a paradigm shift. When discussing the ongoing developments, Purie noted, “Sana does not take away from the brilliance of real-life anchors, who will be mentoring her.” Purie also said that this partnership is likely to reveal creative magic and that there will never be a competition between humans and AI in it. Russia’s Svoye TV, which introduced virtual weather presenter Snezhana Tumanova, assured that it was some way from replacing her human colleagues with bots.
However, what would happen if these AI anchors matured, as generative AI does? And what if they eventually outsmart human anchors? Could that just be explained by AI eating up some more jobs, or could that trigger a dangerous democratic breakdown in ways not imagined at the present moment? Kalli Purie said Sana was completely under her “control.” But what happens if a bot escapes “control”? Nobody knows, possibly. But dystopia of the scifi kind would haunt us.
How much an AI bot might expand its capacity is another grave concern. While bots are capable of reading news reports, we would like to think that they won’t be able to react and engage in arguments while conducting debates the way that individual anchors do now. Can an AI anchor coordinate a debate as effectively as the best human anchors? Undoubtedly, a human’s thought process differs greatly from that of an AI. However, as generative AI develops, could an AI newsreader soon offer the “human touch”? Could Christiane Amanpour, Stephen Sackur, or Fiona Bruce be replaced by an AI anchor? Really?
(The writer is Professor of Statistics, Indian Statistical Institute, Kolkata.)