New research shows most companies lack the policies needed to reassure consumers as they adopt agentic AI in customer experience strategies.
New data from Genesys reveals a critical disconnect between how companies govern AI and what consumers need to feel safe using it.
Four out of five consumers surveyed said they want clear governance of AI interactions, yet only 31% of business leaders reported having comprehensive organisation-wide AI policies and oversight in place.
As agentic AI – autonomous systems that can think, act and make decisions independently – becomes more common in enterprise customer experience (CX) strategies, 91% of CX leaders believe it will enable faster, more effective and personalised service.
However, the findings highlight a complex reality. While enthusiasm is high, governance structures lag behind, posing risks to trust, brand reputation and regulatory compliance.
“Agentic AI is opening up exciting new possibilities for how organisations serve their customers, but earning consumer trust has to grow alongside that progress,” said Olivier Jouve, Chief Product Officer at Genesys. “As these systems take on more responsibility, it’s essential that businesses stay transparent and accountable in how they’re used. With the right guardrails in place from the start, companies can build lasting confidence by responsibly innovating customer experiences that deliver new levels of personalisation and effectiveness.”
Governance gap undermines trust
More than 90% of CX leaders agree strong governance is critical to protect brand reputation (91%), build long-term customer trust and loyalty (91%) and increase consumer comfort with autonomous systems (90%).
Yet over one-third (35%) admit they have little to no formal governance policies. Even more concerning, 28% of those with no policy at all still believe they are ready to deploy agentic AI.
Consumers remain wary, largely due to a lack of transparency on how their data is used and the absence of clear oversight. The top concern was clarity on how AI uses personal data.
Mistrust is compounded by fears of AI “hallucinations”. While 37% of consumers believe AI fabricates information, 59% of CX leaders acknowledge hallucinations risk damaging customer loyalty, increasing litigation exposure and harming brand reputation.
The gap between what leaders know is necessary and what is implemented is troubling given the clear demand for transparency and oversight. Businesses must address these shortcomings before deploying agentic AI at scale.
Consumer trust remains fragile
While 81% of CX leaders trust agentic AI with sensitive customer data, only 36% of consumers do. The disconnect is sharper in high-stakes scenarios.
Seventy-four percent of businesses are comfortable using agentic AI for billing, financial transactions and account security. By contrast, only 35% of consumers are comfortable with AI handling money transfers, 49% with resolving billing issues and 50% with updating personal details.
Still, the research points to opportunity. Over half of consumers (58%) say they do not care whether their issue is resolved by a human or AI, as long as it is handled quickly and completely.
This suggests efficiency and effectiveness can overcome scepticism, but only when paired with transparency and accountability.
Genesys worked with an independent research firm to survey 4,000 consumers and 1,600 CX and IT decision-makers in more than 10 countries in June 2025. Industries represented included airlines, automotive, banking, government, healthcare, insurance, manufacturing, media and entertainment, professional services, retail, travel and hospitality, technology, telecommunications and utilities.
