San Francisco, Aug. 7, 2025 — New knowledge launched at the moment by Genesys uncovers a disconnect between how corporations govern AI and what customers have to really feel secure utilizing it. Whereas 4 out of 5 customers surveyed mentioned they need clear governance of AI interactions, lower than a 3rd (31 p.c) of enterprise leaders say their organizations have complete, organization-wide AI insurance policies and oversight in place.
As agentic AI — autonomous programs that may suppose, act and make selections independently — turns into extra pervasive in enterprise buyer expertise (CX) methods, 91 p.c of CX leaders surveyed consider it should empower their organizations to ship sooner, simpler, customized service. Nevertheless, the information reveals a posh actuality: Whereas enthusiasm is excessive for the transformative worth agentic AI can ship, governance constructions are lagging, posing dangers to shopper belief, model fame and regulatory compliance.
“Agentic AI is opening up thrilling new potentialities for a way organizations serve their clients, however incomes shopper belief has to develop alongside that progress,” mentioned Olivier Jouve, chief product officer at Genesys, a cloud AI-powered expertise orchestration firm. “As these programs tackle extra accountability, it’s important that companies keep clear and accountable in how they’re used. With the fitting guardrails in place from the beginning, corporations can construct lasting confidence by responsibly innovating buyer experiences that ship new ranges of personalization and effectiveness.”
Greater than 90 p.c of CX leaders surveyed agree that robust governance is crucial to: shield model fame (91%), construct long-term belief and loyalty with clients (91%), and improve shopper consolation with autonomous programs (90%).
But many organizations are nonetheless unprepared. Over one-third of CX leaders (35 p.c) admit they’ve little to no formal governance insurance policies in place. Much more regarding, 28% of these with none coverage in any respect nonetheless consider their organizations are able to deploy agentic AI.
This governance hole turns into extra troubling when considered alongside customers’ issues. Many stay cautious attributable to a scarcity of transparency round how their knowledge is used and the absence of clear oversight.
In actual fact, readability on how AI makes use of their private knowledge was the primary concern amongst respondents. Compounding this, 37 p.c of customers consider AI “hallucinates” or fabricates data — a view echoed by 59 p.c of CX leaders, who acknowledge that hallucinations pose severe dangers to buyer loyalty, litigation and model fame. This perceived lack of reliability additional undermines belief and reinforces the necessity for safeguards that guarantee accuracy and accountability in AI-driven experiences.
The hole between what leaders know is important and what organizations have applied is particularly regarding given the clear demand from customers for transparency and oversight. This makes it crucial for companies to shut the hole earlier than deploying agentic AI at scale.
Whereas 81 p.c of CX leaders belief agentic AI with delicate buyer knowledge, solely 36% of customers really feel the identical. This disconnect is sharper in larger stakes conditions.
Companies present robust confidence in utilizing agentic AI for crucial buyer capabilities: 74 p.c of companies say they’re snug utilizing the expertise for billing, monetary transactions and account safety. Shoppers, nonetheless, are far more hesitant: Solely 35 p.c are snug with agentic AI dealing with cash transfers, 49 p.c with resolving billing points and 50 p.c with updating private data.
Nonetheless, the information reveals an essential alternative. Greater than half of customers (58 p.c) say they don’t care whether or not their situation is resolved by a human or AI, so long as its dealt with rapidly and fully.
This alerts that CX effectivity and effectiveness can overcome skepticism, however solely when paired with transparency and accountability. To shut the belief hole and innovate responsibly, organizations should scale in a consumer-centric means
Methodology: Genesys labored with an unbiased analysis agency to survey 4,000 customers and 1,600 CX and IT decision-makers in additional than 10 nations. The survey was carried out in June 2025. Among the many enterprise respondents, the industries represented have been airways, automotive, banking, authorities, healthcare, insurance coverage, manufacturing, media and leisure, skilled companies, retail, journey and hospitality, expertise, telecommunications, and utilities.
