Virtual influencers still lose the credibility race. That is the rough takeaway from experiments run by Myntra and other retail brands testing AI-generated personas in their marketing stacks. The headline finding, as reported by Agency Reporter: synthetic avatars cannot match the trust a real person earns from a buyer.

The wider lens matters. Myntra is one of several platforms that have spent the past two years plugging virtual influencer characters into live campaigns — digital faces backed by machine-learning models that produce captions, photos, and short videos at scale. The pitch was familiar: cheaper content, no scheduling drama, no personal-scandal risk.

The 2024–2025 record is less flattering.

What did the experiments actually test?

Brands swapped virtual influencers into live campaigns and measured the response in different ways. The signals ranged from survey-based trust scores to engagement and repeat-purchase data. Each test repeated the same pattern in its own way: reach from the avatar, belief from the human.

A buyer scrolling past an AI face might pause for the image. Follow-through into the cart — the click that turns to commitment — stayed stubbornly lower for the synthetic creator. Top-of-funnel attention arrived. Bottom-of-funnel trust did not.

Why does the gap stay open?

Virtual influencers look polished. They also feel frictionless in a way consumers have started to read as a tell. Audiences trained on TikTok and Instagram can spot artificial origins within seconds. A face without blemish, a caption without typo, a reply without delay — each marker becomes a reason to discount the message.

Human creators bring texture. They show up tired, get facts wrong, pick fights in the comments. That mess reads as authentic. Synthetic personas cannot fake it convincingly without sliding into the uncanny valley (the creepiness threshold where near-perfect human imitations feel unsettling) of believability.

What's the cost trade-off?

Virtual influencers were pitched as a way to compress production budgets. A single licensed avatar can serve dozens of campaigns without travel, wardrobe, or per-post fees. The marginal cost of an additional asset drops sharply once the model is built.

That savings does not vanish in the trust gap. But it forces a sharper question for the marketing operations team: at what spend threshold does a virtual creator stop being a bargain? Brands running the math now weigh the cost of avatar licensing plus the extra human-creator spend needed to recover the lost trust.

Where does this leave retail marketing teams?

Brands that invested early now face a choice. Keep running virtual influencers as reach tools, where raw impressions matter more than belief. Or pair the avatars with human creators for the trust-heavy work — recommendations, reviews, purchase nudges.

Most marketers landing on the second path treat the virtual influencer as production staff, not a relationship. The avatar shoots the imagery. The human delivers the verdict.

The takeaway

Virtual influencers work as a budget lever. They do not, yet, work as a credibility substitute. The data keeps telling brands the same thing.