Last November, RRR director S.S. Rajamouli gathered the international press and about 50,000 of his biggest fans in Hyderabad, India to unveil his next film, Varanasi. That title was accompanied by a three-minute video that could best be described as animated concept art, a mix of live-action and CGI used to tease the film and provide a glimpse at the director’s epic vision.

The response, both at the event and online, was near-universal praise. However, one small but vocal contingent raised an increasingly common concern: Was the Varanasi teaser made using generative AI?

So when I sat down to interview Rajamouli a few days later, I posed the question to him directly. Short answer? No. Long answer: extremely no.

“It was October when my artists started drawing lines of the teaser — for a three-minute teaser,” Rajamouli says. “And from then until the last day before we released it, the artists were working relentlessly day and night. For the last three weeks, I don't know many artists had a good night's sleep while working on this teaser. And to say that it's AI generated would be a huge insult to their artistic genius. Every single frame took 20 hours to render. That's how hard they worked on it.”

However, while the director says he didn’t use generative AI to create Varanasi’s visuals, he says he does see some benefits from the technology, calling it a “great assistant.”

“It speeds up the processes when you are at a very basic level to give you some ideas,” Rajamouli says, “but at present, I think that's it. Apart from that, it is just one tool that you have to help you — one more assistant. I have a number of assistants while working who help me do my job. So AI is also an assistant. A good assistant at the max.”