AI trust is rising, but corporate AI messaging is losing attention
CARMA finds public trust in AI rising even as media grows more sceptical and audiences tune out vague corporate AI claims.
CARMA has found a counterintuitive shift in the AI reputation story: public trust is moving up at the same time media coverage is becoming more sceptical. For brands and communications teams, that means a more critical press environment does not automatically translate into a more distrustful audience.
The more important problem may be messaging quality. Marketing-Interactive reports that CARMA's latest Perceptions of AI research shows audiences are getting more comfortable with AI while becoming less patient with vague corporate claims about it. The lesson is not that people suddenly trust AI without reservations. It is that familiarity is rising faster than tolerance for buzzwords.
Key Takeaways
- CARMA found public trust in AI rose while media trust signals moved in the opposite direction.
- Audiences increasingly associate AI trust with transparency about how systems work and how decisions are made.
- For brands, generic AI positioning is becoming a credibility risk as audiences tune out exaggerated or vague claims.
Table of contents
Jump to each section:
- Public trust and media trust are moving apart
- The real trust signal is transparency
- AI hype is becoming a messaging liability
- Why brands should stop treating media tone as a proxy for public sentiment
- What communications teams should change now
Public trust and media trust are moving apart
CARMA's finding is useful because it breaks a common assumption in reputation work: that more negative coverage should mean less public trust. In this dataset, the two moved in opposite directions.
Public trust in AI rose from 26% in 2025 to 30% in 2026, while media trust signals fell from 35% to 26%. These figures come from CARMA's latest study, as reported by Marketing-Interactive.
The study compared media analysis with audience research rather than treating one as a stand-in for the other. CARMA's Perceptions of AI: A Power Shift examined thousands of tier-one media articles across dozens of markets and paired that with consumer research across multiple countries. That matters because the gap between press narratives and audience perceptions is the point, not a statistical quirk to smooth over.
Trust also became a more prominent public emotion than excitement. In other words, people may be moving past the novelty phase of AI. As the technology becomes normal, audiences can become simultaneously more trusting and more demanding.
The real trust signal is transparency
One of the clearest differences in CARMA's findings is what audiences and media appear to emphasize when they talk about trust. Media coverage puts more weight on safety, reliability and accountability. Audiences, meanwhile, are more likely to connect trust with understanding how AI systems operate and how decisions are made.
That is a meaningful distinction for marketers. A brand can publish a long list of safety commitments and still leave people unsure about what the AI actually does, what data it uses or where human judgment sits in the process.
Jennifer Sanchis, CARMA's Insights & Consulting Director for Europe and the US, framed the communications opportunity in the firm's own commentary: people want to understand how AI systems operate and how they are used to make decisions. Her point is practical. Trust-building content has to explain the machinery, not just praise the outcome.
For communications teams, this creates a higher bar than adding an "AI-powered" label to an existing product story. It means explaining inputs, decision logic at a useful level, human oversight, exceptions and accountability in language a non-technical audience can follow.
AI hype is becoming a messaging liability
The sharpest warning in the research is not about fear of AI itself. It is about disbelief in how companies talk about AI.
59% of respondents said companies exaggerate their use of AI and rely on buzzwords, while 44% said they pay less attention to messages that mention AI. CARMA's findings were reported by Marketing-Interactive.
That should worry brands more than a temporary spike in negative press. If a large share of an audience is predisposed to tune out when AI appears in the message, the problem is not simply sentiment. It is attention.
This changes the value of AI positioning. Calling a product AI-powered can now trigger scepticism unless it is immediately followed by something concrete: what improves, what changes for the user, what data is involved, what the limitations are and where humans remain responsible.
The reputational risk is especially obvious when AI claims overlap with workforce changes. CARMA's commentary notes that coverage becomes notably more negative when executive AI statements appear alongside layoffs or restructuring. In those situations, upbeat productivity messaging can sound less like innovation and more like evasion.
Why brands should stop treating media tone as a proxy for public sentiment
The divergence between media trust signals and public trust is a reminder that reputation teams need separate measures for separate audiences. Media analysis remains important, but it cannot tell a company exactly what customers or the broader public believe.
A more sceptical press may be asking harder questions because AI is becoming more consequential. That scrutiny can coexist with rising public comfort as people use AI more often in work, search, entertainment and everyday software.
The opposite can also happen. Positive coverage does not guarantee trust if the public sees corporate messaging as inflated. This is where communicators can get misled by dashboards that collapse sentiment, share of voice and trust into one composite narrative.
The better approach is to read media tone and audience trust as related but distinct signals. One tells you how the public conversation is being framed. The other tells you whether people accept, question or reject the claims being made.
What communications teams should change now
For brands, the practical response is not to communicate less about AI. It is to communicate with more evidence and less theatre.
Start with the specific job the system performs. Explain what has changed for the customer or employee. Be explicit about where data comes from, how outputs are reviewed and what remains under human control. If the system has limitations, name them before critics do.
This is also a reason to diversify spokespeople. CARMA's analysis has found that CEOs generate some of the most positive AI coverage, but academics are more trusted on the topic. Executive visibility still matters, but credibility may improve when technical experts, independent researchers or operational leaders can explain how the system actually works.
The broader shift is that AI communications are entering a maturity phase. Audiences are becoming more familiar with the technology, but familiarity is removing the benefit of the doubt. Brands can no longer assume that more AI language creates more confidence.
Public trust may be rising. The burden on communicators is rising with it.
