Havas published a September 2026 global study, Avoiding the AI Efficiency Trap, on how consumers rate AI-enabled brand experiences. Across more than 21,000 scenario evaluations, only about 9% landed in the highest Valuable AI Experience Score band (81–100). High-scoring experiences were linked to 4.4× higher engagement and 1.6× stronger brand desire and growth potential than low-scoring ones. Those multipliers are Havas’s research claims.
The fieldwork paired Juno’s 200 in-depth interviews with YouGov quantitative work among 10,558 respondents in seven markets: Brazil, China, France, Germany, India, the UK and the USA. Five sectors and seven AI touchpoints were scored. This is a global panel, not a Jamaica-only survey. The pattern still lands here, where WhatsApp Business bots, chat support and AI product copy are already in the mix.
Havas names the gap the AI Efficiency Trap: AI that saves the brand money or time without creating enough value for the customer. Across the seven markets, 56% said AI used by brands can feel misleading, 50% said brand AI can make things look, feel or sound the same, and only 29% said brand AI feels human. Trust to use AI responsibly sat at 35%. Sixty-seven percent agreed AI customer service can frustrate when it leaves people stuck in a loop. Seventy-eight percent said clear disclosure when AI is used matters.
The Valuable AI Experience Score weighs Utility, Credibility and Appeal, and subtracts for experiences that feel misleading or manipulative. AI performed best when it helped people do a job: education and support scored 56.7 out of 100, product comparison 55.4, and brand presence in chat tools 54.9. AI-generated influencer content scored lowest at 45.9, followed by brand advertising at 48.4 and shoppable formats at 49.6. Automotive led among sectors; healthcare OTC trailed.
What this means. For Jamaican and Caribbean brands rolling out chatbots, WhatsApp agents or AI product helpers, Havas’s read is blunt: efficiency for the company is not the same as value for the customer. Help people compare, decide or get support, and disclose when AI is in the loop. More branded AI content without a clear job risks sameness and distrust. Treat the percentages as Havas’s seven-market findings, not STATIN data.
What to do. Audit every live AI touchpoint this week — WhatsApp bot, site chat, product recommender, AI-written ads — and ask whether it helps the customer finish a task or only cuts your headcount. Prefer education, comparison and support over AI influencer filler. Put a plain disclosure where AI answers. If one change fits this week, kill the loop that strands customers without a human handoff. Last fact-checked 24 September 2026.
