Keep an ‘AI’ on the news 👁️👁️
Good afternoon!
Welcome to The Impression, your weekly primer on the business of media, entertainment, and content.
If someone shared this newsletter with you or if you’ve found the online version, hit the button below to subscribe now—it’s free! You can unsubscribe anytime.
In the 1980s, the few Indian homes that had a TV almost always tuned into the news on Doordarshan, the country’s only broadcaster. Those bulletins were formal, gentle affairs. Among the iconic anchors of this pre-privatisation era was Salma Sultan, famous for her soft voice and the rose always tucked in her hair.
Today, things are a lot different. There are more TV channels than we know what to do with. News bulletins have been replaced with nightly “debates”. Fiction often enters the newsroom. But one fixture remains: the anchor.
You may long for the staid anchors of the ‘80s. Or perhaps you prefer the flashier ones that started dominating news channels from the 2000s.
But nowadays, there’s a brand-new type of news anchor that’s getting popular: one generated by AI.
Can AI anchors fix TV news?
India Today/YouTube
At last week’s India Today Conclave in Mumbai, India Today Group’s vice chairperson Kalli Purie made an announcement.
“In March this year, we added the newest member of our team,” she said. “The first one that we did not hire in 48 years. We created from scratch our first AI anchor, Sana. Sana has done over 200 hours of programming across different genres, languages, and platforms.”
Sana, the AI anchor, then introduced India Today’s “AI family” — five AI-generated anchors speaking five languages, including English. Fun fact: all their names have ‘AI’ in them – Naina, Aina, Aishwarya, Saili and Jai.
Since March this year, the India Today Group’s Hindi news channel Aaj Tak has been heavily experimenting with AI news presenters. These anchors are interviewing film stars and presenting news bulletins in various Indian languages. The YouTube channel Aajtak AI has over 4,500 subscribers and almost 700,000 views since March this year.
Then, last month, Aaj Tak debuted an AI avatar of its star Hindi anchor, Anjana Om Kashyap.
Aaj Tak isn’t alone. In July, Kannada channel Power TV launched an AI anchor named Soundarya, while Odia TV channel OTV introduced Lisa.
Why are the India Today Group and other news networks introducing AI-generated hosts?
Kashyap has one answer. “Elections are coming. I will be busy field reporting from various constituencies, interviewing people. I will need a partner just like me. Now, I can do Halla Bol, Special Report, and many other shows all at once.”
Although it’s very early days, AI may help star anchors do everything everywhere all at once. But are anchors, and journalists aspiring to be one, ready for direct competition with AI?
Legal Loopholes
Under current Indian laws, media professionals, including actors, writers, and journalists, may not have adequate protection from being replaced by AI-generated doppelgangers, according to Ameet Datta, intellectual property lawyer and senior partner at law firm Saikrishna & Associates.
“India does not have any majority judgement or statutory provision to support the right to publicity,” he told The Impression. The right to publicity is an individual’s right to protect their personal image and brand and make money off it. Indian laws do not explicitly identify this as a right. At best, our courts accept legal protection from “passing off” under trademark laws, Datta says. “You can’t ride on someone else’s reputation. So, for example, using Amitabh Bachchan’s name and likeness to sell a pen or a calculator violates the law because you’re ‘passing off’ his status as an endorsement for the product.”
What this means: if you have used your name or likeness for commercial purposes, say to endorse a brand, you may have a case for protecting it. That, too, comes with caveats. Copyright laws have provisions that allow people to use protected material under ‘fair use’, such as journalism and satire.
Actors Anil Kapoor and Amitabh Bachchan have also secured copyright protection of their name, image, voice, and likeness — they cannot be used without consent in ads or other commercial pursuits but can be used for parody or satire. That’s where AI also comes into the picture.
Take Anil Kapoor, for example. Last month, the Delhi High Court granted legal protection to his name and even his catchphrase “jhakaas” (Marathi slang for ‘awesome’)—not for the word outright, but only when said in his signature style. Justice Pratibha Singh also banned the use of AI to use Kapoor’s image to make GIFs and other common media for monetary gains. She noted that AI-generated content, such as deepfakes, can hurt a celebrity’s reputation and violate their right to privacy, which is a fundamental right.
But getting legal protection for what you are famous for isn’t so straightforward.
“Actors like Anil Kapoor ask to protect their likeness in a movie character, like Mr India,” Datta says. “But do they own the copyright of their appearance and character in that film? It could be argued that this is the property of the film’s producer, not the actor.”
How might these deficiencies in the law affect news anchors? To prevent a news channel (or anyone else) from using their AI avatar, they will also have to seek legal protection from the courts for specific aspects of their likeness that are signature to them. Only actors have received such protection so far because they use their names and faces for endorsements. News anchors may need to come up with a new set of arguments for why they deserve similar protection.
Moreover, many news anchors may have signed away the rights to their likeness to their employers or may not have a clause regarding this in their contracts. Generally, media companies claim copyright ownership of all their journalists’ work. Wouldn’t videos of an anchor hosting a news show also qualify? A channel could legally use this material to create AI copies of their anchors or train their AI models to generate more members of their ‘AI family’ of anchors.
Kalli Purie and Anjana Om Kashyap did not respond to my requests for comments.
It’s ironic to watch Kashyap excitedly introduce her own AI avatar to take over the work she is too busy to do. Compare this with the striking actors of Hollywood’s SAG-AFTRA union. Among their demands was that studios must not scan the likeness of background actors to be used forever without consent or future compensation. Background actors in big-budget productions like Disney’s The Mandalorian said they felt violated after being scanned on sets and never getting called for work again.
Not There Yet
Star anchors may be able to hold on to their jobs and audience through the rise of AI. The bigger benefit of AI-generated anchors could lie in replacing human hires forever. But for now, anchors aren’t too worried.
“It’s early days. Of the AI anchors I’ve seen and heard so far, they’re still not quite there,” Mangalam Maloo, assistant editor and anchor with CNBC-TV18, told The Impression. “One will have to wait for more iterations of the same before they’re actually on-air ready. On a lighter note, given the current salaries in media, the amount of investment it will take to develop that AI anchor now, we’ll get many real anchors at a much lower cost.”
Maloo also added that AI anchors could take over “commoditised work”, leaving editors and news producers more time to focus on more onerous projects in the newsroom.
Tejasvita Singh, an anchor on CNN-News18, agrees. “One window for using an AI anchor is in news bulletins that are broadcast late at night, when we don’t have live news,” she told The Impression. “In these programmes, an anchor usually reads out pre-written story summaries from a teleprompter. That is a job an AI anchor could do well.”
Singh points out another use case for AI anchors: “fake lives”. News channels are known to record an interview with a guest, but edit it for a primetime programme so that it appears to be a live interview. The anchor simply asks questions again and snippets of the interview are played in response. Instead of a human anchor, an AI replacement could repeat these questions when the interview goes “live”, Singh says. Writer Devdutt Pattanaik recently took issue with this practice when India Today played an edited version of his pre-recorded interview as a debate with economist Sanjeev Sanyal, who was live.
However, some journalists just starting out as anchors feel the role and visibility of the profession is slowly changing across newsrooms. “This [news anchoring] is already a depleted role,” Singh says. “In many news programmes now, the anchor’s voice may be heard but the camera isn’t always focused on them. News studios sometimes use such zoomed-out angles that the anchor is barely visible. In those instances, an ordinary viewer may not be able to tell an AI anchor apart from a human one.”
Despite the doppelgangers, star anchors may not be as threatened by replacement with AI. But mid- and early-career anchors who are trying to cultivate a public image and recognition may find themselves redundant if they are relegated to programmes in odd time slots.
Maloo also says smart reporters who know how to add value in a news report will be able to hold the threat off.
“Just a little native intelligence and the willingness to dig deeper will keep one ahead of AI anchors. AI will first take the generalists away. Natural foolishness is a bigger threat than artificial intelligence.”
Mangalam Maloo, assistant editor, anchor, CNBC-TV18
All this comes with the assumption that channels won’t misuse AI anchors or avatars of existing hosts. What if your AI avatar read out news in a derogatory manner, or worse, said something on-air that goes against the country’s laws? According to Datta, journalists and news channels aren’t even considering such possibilities, let alone planning for them.
According to reporters I spoke to on- and off-record, TV journalists aren’t worrying about AI anchors just yet. There are bigger problems plaguing the news business. The news genre isn’t growing even as TV viewership declines among younger, urban, affluent audiences, as per data from the FICCI-EY Media & Entertainment report 2023 (pdf). Besides, ad revenues for TV channels are also stagnating, while several large consumer brands are keeping away from them altogether, citing ‘toxic’ content and an ideological slant in their coverage.
News channels need a new gimmick to grab attention. AI anchors come with an upfront investment. But in the long run, they could help legacy media companies grab eyeballs while cutting costs. In the absence of formal unions or explicit legal protections, there is little that Indian newsroom employees will be able to do to prevent all of this. Some may benefit, but many may see their roles reduced or simply replaced.
How many years will it be before someone resurrects a young Salma Sultan to read us the news again, a fresh rose tucked in her hair?
Last Scroll Down📲
Tenor
Pay and Skip: Meta may launch ad-free versions of Facebook and Instagram in India for a subscription plan next year, Livemint reported. It’s already planning an ad-free pilot in Europe for €10/month to compensate for the revenue it may lose from regulations curbing personalised ads. Marketing agencies and brands that heavily rely on digital advertising say they may have to start thinking about their next steps. Although, given India is known for a high-volume, low-average revenue user base, nobody needs to panic just yet.
What were they thinking?: Meanwhile, Meta rolled out AI stickers across its family of apps late last month. But it looks like no one thought to place restrictions on the feature. Facebook users are reportedly using AI text-to-image prompts to generate irreverent (and rather lewd) stickers to share in direct messages. Specimens include various children’s cartoon characters armed with assault rifles or heavily pregnant and well-endowed versions of Elon Musk and Karl Marx 😵
Suitors line up: Disney is knocking on all doors, Star India in hand. PE firm Blackrock is reportedly in talks to buy out Disney’s India operations, including TV and the streaming platform Hotstar. Earlier, Bloomberg reported Disney officials were in talks with Mukesh Ambani’s Reliance Industries, as well as the Adani Group and Sun TV.
Holding off the Yanks: European networks are uniting to hold off competition from Netflix and Amazon Prime. Eight publicly funded broadcasters have formed a ‘New8’ pact to produce high-end TV shows and distribute them worldwide. New8 will start with eight originals in the next three years. Several Netflix originals made in Europe have become big international hits, including Money Heist (Spain), Dark (Germany), and Lupin (France).
Trumpet 🎺
@profcgoldin/X
It is now stupidly easy to spread fake news on X/Twitter. On October 10, leading news organisations reported that Nobel Prize-winning economist Amartya Sen had died. All reports led to one post, seemingly from the official account of Claudia Goldin, who just won the economics Nobel herself.
Turns out, the account was fake (this is her real account, inactive for many years now). Meanwhile, the fake news poster claimed to be a hoax created by Italian school teacher Tommaso De Benedetti, although this hasn’t been verified either.
@profcgoldin/X
More than a decade ago, De Benedetti became famous for tweeting fake news (such as the death of the Pope) and impersonating world leaders (such as Syrian president Bashar al-Assad). Before that, he fooled Italian newspapers into publishing fake interviews with famous American authors. This 2012 piece in The Guardian details his exploits.
De Benedetti claimed he does it to expose “weak” news media and demonstrate how unreliable social media can be.
The situation has only worsened in the last 10 years.
Now that anyone can buy verification on X, it’s exponentially easier to pass off fake news for real. The Goldin impersonator didn’t even use a blue tick; all it had to do to be taken seriously was add “official account” to its bio. The Times of India (archive link), Zee News (archive link), and Deccan Herald (archive link), among many others, reproduced the news without fact-checking it.
Meanwhile, the fake Goldin has deleted all their tweets and changed their name to ‘Gatto di Pino’ with a profile picture of a tuxedo cat. Perhaps it is an Italian reference to Figaro, the cat who belonged to Pinocchio: the puppet whose nose got longer every time he lied.
That’s all this week. If you enjoyed reading The Impression, please share it with your friends, family, and colleagues. And please write to me anytime at soumya@thesignal.co with thoughts, feedback, criticism or anything you’d like to see discussed in this space. I'd love to hear from you.
Thanks for reading, and see you again next Wednesday!