The News
New research from Idea Grove finds that while AI tools like ChatGPT and Google Gemini are rapidly becoming part of how consumers discover brands, they are not yet trusted to drive purchase decisions. In a survey of 1,000 U.S. consumers, only 2% said they would buy from an AI-recommended brand without verifying it first, while 98% conduct additional research through reviews, search, or other sources.
The study highlights a growing gap between AI-driven discovery and consumer trust, with traditional validation signals, such as customer reviews, search rankings, and press coverage, continuing to play a decisive role in purchase behavior.
Analysis
AI Is Reshaping Discovery, Not Replacing Decision-Making
One of the clearest takeaways from this research is that AI is changing where the buying journey starts, but not where it ends. According to the study, 42% of consumers now use ChatGPT for brand research, and 38% use Google Gemini. That level of adoption is significant, especially considering how quickly these tools have entered the mainstream.
Adoption, however, does not equal trust. Only 15% of consumers say they fully trust AI-generated recommendations, and 85% express some level of skepticism. From an application development perspective, this reflects a broader pattern. While AI is accelerating workflows and surfacing information faster, it still depends on underlying systems of record and validation to produce reliable outcomes.
In this case, AI helps users discover options but it doesn’t replace the need to verify them.
Trust Still Lives Outside the AI Interface
What consumers do after receiving an AI recommendation is where things get more interesting. The data shows that 45% of users immediately search for the brand on Google, while others turn to review platforms or the company’s website. Only 2% skip that step entirely. This reinforces something that shows up consistently in application development research: speed is increasing, but confidence still requires validation.
Efficiently Connected’s AppDev 2025 research shows that 46.5% of organizations are under pressure to deliver applications 50–100% faster than three years ago. That same pressure exists on the consumer side. People want faster answers, but they still need to confirm those answers before acting on them.
In practice, that creates a two-layer model where AI handles discovery and summarization and external systems (search, reviews, websites) handle validation and trust
Market Challenges and Insights
There’s also a disconnect forming between how brands are investing and how consumers actually behave. A growing ecosystem is forming around optimizing for AI recommendations, essentially trying to ensure that a brand shows up when users query tools like ChatGPT. But this research suggests that visibility alone isn’t enough.
Consumers are still relying on familiar trust signals. According to the survey:
- 78% say customer reviews increase purchase confidence
- 71% look at search rankings
- 69% consider how long a business has existed
- 58% value press coverage
Even more telling, 69% of consumers said they would choose a brand with press coverage over one with no visible presence, even if both were recommended by AI.
At the same time, awareness of how AI recommendations work is still limited. Nearly half of consumers don’t realize that companies actively try to influence AI outputs. This creates an uneven landscape where trust decisions are being made without full visibility into how recommendations are generated.
AI and Trust Signals Are Converging
One of the more subtle implications of this research is that AI systems and human decision-making are relying on the same underlying signals. AI models are trained on content like reviews, articles, and web data, and these are the same sources consumers turn to when verifying a brand. That creates a feedback loop where:
- Strong trust signals improve AI visibility
- AI visibility increases initial discovery
- But trust signals still determine final decisions
Paul Nashawaty’s research often frames this as a shift toward connected ecosystems, where different layers of the application stack work together rather than independently. For developers and platform teams, this is a useful analogy. AI doesn’t replace existing systems; it sits on top of them and depends on them.
Looking Ahead
As AI becomes more embedded in search and discovery, the gap between visibility and trust will likely become more important, not less. This research suggests that AI will continue to grow as an entry point for decision-making, but traditional signals will remain critical for closing the loop.
For developers and platform builders, this has broader implications. Systems that combine AI-driven interfaces with strong validation layers will likely be more effective than those that rely on AI alone. It also raises questions about how platforms present and prioritize trust signals within AI-generated experiences. This is something that will likely evolve quickly as both user expectations and regulatory scrutiny increase.
TinyMCE AI Embeds Writing Intelligence Directly Into Dev Workflows
Twilio Expands Flex With Embeddable Contact Center Architecture
Telestream Expands Multi-Cloud Media Workflows with Oracle Cloud
New Research Finds AI Recommendations Drive Discovery, But Not Decisions
Mirantis and NVIDIA Automate AI Factory Deployments at Scale
