Marketing in a Trust Deficit

Executive Summary Generative AI has cut the cost of producing convincing, false content to near zero, turning brand safety from a marketing task into a shared mandate across legal, IT and the CEO's office.

Over-reliance on AI-assisted research has weakened marketers' direct contact with customers; organisations that retain judgement and thinking in-house are better placed to withstand scrutiny than those that outsource it.

Every device owner is now a potential publisher, and brands face greater risk from seeded misinformation and competitor-driven defamation than from organic criticism , which puts monitoring on a level with campaign planning.

Rather than being a differentiator, transparency has become a baseline expectation. What now earns consumer goodwill is visible restraint, such as a refusal to use AI in creative work , rather than messaging alone.

Discovery is shifting from search results to answer engines, making a brand's presence in AI-generated summaries as strategically important as its presence in traditional media.

. Over the last decade, marketing teams built their practices around cheap reach and increasingly precise personalisation. Generative AI has begun to erode both foundations at once: producing content now costs next to nothing, deepfakes are indistinguishable from the genuine article, and the platforms marketers depend on to reach consumers reward outrage over accuracy. For CMOs, brand safety and reputation management have moved from a downstream task, handled after the campaign went live, to a governance question that the legal, IT and CEO offices now have a stake in. At a recent India CMO Forum open house, chaired by Amit Sinha Roy, Executive Director at CII Digital, our members examined how far things have shifted and what rebuilding consumer trust may require of the marketing function.

The Collapse of Cheap, Convincing Content GenAI has removed the two frictions that once forced misinformation to travel slowly: cost and craft. AI can now easily convert documents (such as equipment manuals) into instructional videos, and even turn non-visual content (such as dry, technical audio) into visuals that appeal to a whole new audience. The toolsets that have made this transformation possible also makes it just as easy to fabricate content, and brands often discover the difference only after it has already spread.

Governance systems are being forced to catch up, and legal and tech team are becoming – or soon will – key stakeholders in the content/marketing dissemination process. This is vital, given that pressures to expand AI's use often comes from above. Increasingly, Marketing feels pushed by senior leadership into generative AI use-cases without full commitment to enterprise-grade tools, and without a clear grasp of the privacy exposure this creates. Customer relationship data might pass through unlicensed, consumer-facing systems with no contractual guarantee of confidentiality, a risk few teams fully register. For CMOs on either side of the fence – whether reluctant to adopt AI or pushed to overuse it – the DPDP (Digital Personal Data Protection) Act offers a formal, external basis to hold the line, rather than relying on internal pushback alone. This is as much a cybersecurity exposure as a marketing one, impacting brand safety commitments, whether implicit or explicit, that assume that customer data stays inside the organisation.

The Erosion of Primary Marketing Judgement Access to instant answers has changed what marketers do with their time. For one, the ability to get easy answers online has reduced the pressure to speak with customers directly, visit the ground or investigate a complaint before responding to it. Where an earlier generation of marketers built judgement by carefully checking claims against individual cases, many now default to querying an LLM instead, and often don’t bother to verify what comes back.

There remains, however, a firm line between automating well-defined work and outsourcing the thinking itself. AI should never be used for original research meant for public consumption, only to speed up execution of work that has already been reasoned through internally. The default setting should remain manual, with a human involved at every stage, not just at the point of final review.

This, of course, does not constitute a departure from AI usage in its entirety. Increasingly, discoverability runs through AI-driven engines rather than search tools, and most consumers now treat chatbot responses as a first-line source before diving deeper. Marketers who withdraw entirely from these channels risk ceding the moment of product discovery to a process they cannot influence.

Brand Vulnerability in a Creator Economy Barriers to publishing have fallen for everyone. A smartphone and a strong point of view are now enough to build an audience, flattening the information advantage brands once held. Today, an audience can verify, contradict or reframe a brand's message as quickly as the brand can publish it. No one is exempt from this shift: a competitor planting seeded, defamatory content faces serious risks, given how easily any claim can be published and amplified. Specialist monitoring services exist to solve exactly this problem, tracking brand and product mentions across channels, including parts of the web that standard search engines do not reach. As a result, damaging claims can be identified and addressed before they gather momentum.

Going against the grain, several of our members explained their own decisions not to use AI in creative output. In a world overrun by AI, consumers are more likely to notice non-AI content, often getting attracted to it because of the novelty factor. The same logic extends to investing in a physical presence: as digital content becomes cheaper and less trustworthy, the premium attached to real, unmediated spaces, from flagship stores to in-person events, has only risen.

Trust as the New Discovery Layer Increasingly, consumers discover brands through the summaries AI models produce in response to a search query. Recognising this, one organisation – a diversified corporate group – deliberately built content and public reviews across multiple sources over time. This eventually shaped how AI models answer a direct question about the brand. Other ways of successfully carving out a niche, without having to resort to the usual paid amplification, might include a founder putting out vlog posts about a personal project; postering a single student bar rather than a citywide circuit or; sponsoring a YouTube comedy show with a small but devoted following. The key is to build something distinctive enough that word travels ahead of the campaign.

The main takeaway for marketers is that their role is shifting from persuasion toward earning attention through demonstrated usefulness. Personalisation earns trust only when it visibly serves the customer's interest rather than simply narrowing a sales funnel. Applied carelessly, it can actually erode trust. Younger audiences in particular seem to respond better to being shown a solution than to being reminded of a problem . This shifts the fundamental goal of messaging from being an argument to becoming a demonstration.