The digital economy has often operated on an unspoken promise—the more consumers share, the more relevant their online experiences become. Algorithms have anticipated what people want to watch, buy and read, while brands have celebrated personalisation as a key expression of customer-centricity. Consumers have also welcomed much of that convenience. But as artificial intelligence becomes embedded into everyday digital experiences, people no longer want technology to simply anticipate their needs and behaviours; they increasingly want freedom and control to decide when, where and how it should step in.
That shift came into sharp focus when Meta recently rolled back its AI-powered image generation model which allowed users to generate images using public Instagram accounts. The feature faced immense backlash over privacy concerns and being an automatic opt-in for users, leading to Meta admitting that they ‘missed the mark’.
The episode reflects that convenience is still valued, but not at the expense of agency. Ambika Sharma, Founder & Chief Strategist at Pulp Strategy Communications and Product Architect of NeuroRank, frames this shift clearly, saying, “The feature was not rejected for what it could do. It was rejected for what it assumed. Adult public accounts were included unless someone found the setting and switched it off. The lesson is that consent is an architecture decision made before launch, not a settings page written afterwards.”
As AI reshapes everything from content discovery to advertising, brands are entering an era where earning permission may prove more valuable than simply delivering personalisation.
The AI Conundrum
This shift also raises a more fundamental question about the role consumers want AI to play in their lives. While people readily embrace AI to simplify tasks, surface recommendations or reduce effort, they appear far less comfortable when it starts assuming consent and making decisions for them.
Sharma believes that evidence does not suggest that people are less willing to use AI. “People are using AI more, not less. They ask ChatGPT, Gemini, Claude and Perplexity what to buy and which brand to trust, and they act on the answer. What has changed, however, is narrower and more interesting. People are comfortable with AI deciding what to show them. They are not comfortable with AI deciding what to make of them. The Muse reaction was never about the image model. It was about identity,” she elaborates.
Yasin Hamidani, Director, Media Care Brand Solutions, is also of the opinion that consumers are not rejecting AI, but are increasingly against AI that operates without knowledge or consent. “People appreciate recommendations and automation when they're transparent and genuinely useful. That's why AI disclosures will become increasingly important. In the long run, transparency won't slow adoption; it will strengthen confidence in AI-powered products and marketing,” he says.
Clearly signalling when AI is being used, what role it plays and how consumers can exercise control helps build credibility in an environment where trust is increasingly fragile. Amyn Ghadiali, Country Head, Gozoop Creative, points out that AI disclosure is not just a legal checkbox; it is a brand trust signal. He reveals, “Edelman found that 71 per cent of consumers will lose trust in a brand the moment they discover undisclosed AI use. That is not resistance to AI. That is resistance to being managed without consent. Brands that miss the difference will pay for it in the only currency that matters—loyalty.”
There is another caveat that brand growth consultant Suumit Kapoor highlights. “When Meta rolled back its AI image generation feature, the instinct everywhere was to reach for the same word—disclosure. Label it, flag the AI, and call the problem solved. But the actual objection was never that AI was present; it was that public photos were used to generate new images of people without a real permission layer.”
He highlights that places like Europe and India have, or are rolling out, laws mandating disclosure of synthetic content, while countries like UK, Canada, Japan and Singapore operate on a voluntary principles basis. And yet, he believes that disclosure, as most compliance teams are currently building it—a label, a footer note, a watermark—will satisfy the law in these systems and still fail the actual test Meta failed. “Compliance and trust are being treated as the same project internally at most companies right now. They are not the same project, and the gap between them is where the next Meta-scale backlash is already forming,” he states.
Merely disclosing AI use or AI features is not enough, if it is not accompanied with significant control. “Meta disclosed. There was a setting. People could opt out. It still failed, because the disclosure sat where nobody reads, and the default was on. Disclosure without control is notification. The standard worth adopting has three parts: tell people plainly, set the default to off for anything involving a person's identity, and make reversal a single action,” Sharma says.
Personalised by default or personalised by choice?
If consent and AI disclosures establish the conditions for trust, they also expose a bigger question about how AI-powered experiences should be designed in the first place. Is it enough for brands to inform consumers and offer an opt-out, or should personalisation begin only when users actively opt in?
Ghadiali believes that the industry is already moving towards the latter faster than most brands and agencies are prepared for. “When Apple introduced ATT, 75 to 85 per cent of iOS users opted out of tracking, leaving ad platforms essentially flying blind on the majority of high-value mobile traffic. By Q2 2024, global opt-in rates had fallen to just 13.85 per cent. That is the new world, which is already here. Personalised-by-choice is now a strategic necessity.”
Sharma backs this, adding, “Personalisation by default is inference; it guesses. Personalisation by choice is declaration; someone told you. Declared preference outperforms inferred preference because it removes the guesswork. The brands that ask will hold better data, and a more durable relationship, than the brands that assume.”
Essentially, the era of brands presenting personalisation with the reasoning of ‘because we can’ is giving way to ‘because you said yes’. And in this changing consumer landscape, the value of zero-party data is automatically increasing.
Hamidani elaborates, “Zero-party data is becoming one of the most valuable assets a brand can own because it is voluntarily shared by consumers. It reflects explicit preferences rather than inferred behaviour, making it both more accurate and more compliant with evolving privacy expectations. As third-party signals continue to weaken, brands that invest in collecting and using zero-party data responsibly will have a significant competitive advantage.”
Sharma reveals a second, more underrated shift. “Intent is increasingly declared to AI models rather than to brands. When someone asks Gemini which health insurance suits a family of four in Delhi, that is a statement of intent more precise than any cookie ever captured. The brand does not own that moment—it either appears in the answer or it does not. So, while zero-party data becomes the most valuable data you hold, AI visibility becomes how you earn the chance to collect it.”
Kapoor, however, notes that brands first have to earn enough trust for consumers to share zero-party data, along with clear disclosure of purpose. “Nearly nine in ten loyalty program members say they are willing to share personal information for rewards. The overwhelming majority also say the reason for asking has to be completely obvious first. Willingness is conditional on legibility, not general goodwill, and it varies sharply by category, with high engagement categories like travel and events seeing far more voluntary sharing than lower engagement ones.”
Additionally, according to the PwC India Consumer Data Privacy Survey, many Indian consumers do not know their basic data rights, and a strong majority already suspect their data is not safe with the companies holding it. “That is not a market primed to share freely. It is a market where the request itself, not the AI behind it, is the thing that has to earn trust first,” Kapoor says.
What’s next?
Taken together, these shifts suggest that the future of AI-powered marketing will be defined less by how much data brands can collect and more by how responsibly they collect, and use, it. Permission, transparency and consumer agency are becoming central to how trust is earned and sustained. In India, the DPDP Act is already nudging the ecosystem in that direction.
“Consumers will expect greater visibility into how their data is used, more control over their experiences, and clearer value in exchange for sharing information. Brands that prioritise transparency, consent and relevance will strengthen customer relationships. Those that continue relying on opaque data practices are likely to see declining engagement and lower long-term loyalty,” Yamidani predicts.
Additionally, Ghadiali expects stronger consumer agency to reshape advertising in three irreversible ways. “Attention becomes a premium currency. You earn it or you don't get it. The creative bar rises dramatically when consumers choose what they engage with, so average work gets ignored at scale. And the brand-consumer relationship becomes genuinely bilateral—not a brand speaking at a target through an ad, but an ongoing exchange where the consumer has real power.”
