Seven in ten APAC consumers have walked away from an AI interaction because the system did not know who they were or what they had already said, according to Twilio's 2026 Customer Insights Series, released 17 September 2026. The survey covered 7,652 consumers and 660 business leaders across 18 markets, including Hong Kong, India, Indonesia, Japan, Malaysia, the Philippines and Singapore.

The perception gap between brands and guests is stark. 84% of brands believe their AI recognizes returning customers; 70% of consumers say they often have to start from scratch or repeat information, and the same share has stopped an AI interaction mid-way when context was lost. 65% repeated themselves after an AI-to-human handoff, and 5% ended the brand relationship altogether.

In the Philippines — where Twilio sampled 353 consumers and 35 business leaders between April and May 2026 — the figures run slightly hotter: 73% report repeating themselves, 73% have abandoned AI mid-conversation, and 74% of brands still believe their AI recognizes return customers. One detail deserves operator attention: 46% of Filipino consumers said a fast AI agent without personal context frustrates them more than a slower one that remembers them. Only 16% felt the opposite.

AI already handles 52% of customer-service conversations in APAC, a share Twilio's respondents expect to reach 65% by 2027. Yet only 22% of APAC businesses disclose that customers are talking to AI at the start — against 70% of consumers who want that disclosure up front.

The hotel-specific numbers are worse

A second study, h2c's AI Opportunity Study 2026 — covering 113 hotel chains, more than 8,200 properties, roughly 1.3 million rooms and 230 documented AI implementations — finds AI in use at 91% of chains, up from 78% a year earlier. But only 28% have an enterprise-wide AI strategy, leaving 72% without one, and only 13% report measurable ROI. Chains rate AI's contribution to business performance at 5.6 out of 10.

The adoption pattern shows where the benefits have landed: 67% of chains report better operational efficiency and 59% say AI frees staff for higher-value work, but only 32% cite an improved guest experience. Hotel AI has so far been pointed inward — and 55% of chains now plan guest-facing AI agents, which expands the surface where amnesia shows up.

One data point in the h2c report moves in the wrong direction. Across ten AI use cases tracked year on year, almost everything grew: chatbots in use rose from 42% to 64% of chains, forecasting from 22% to 52%, guest engagement from 29% to 46%. AI-powered personalization fell from 22% to 16% — the only decline — even as 59% of chains say they plan to deploy it. Hotel AI is talking to guests more and remembering them less.

The write-side failure

The most consequential hotel number is 69%: the share of chains that still enter or enrich guest preference data by hand. Only 19% use AI to extract preferences from guest interactions, and another 19% do not actively enhance preference data at all. Only 42% of chains have a central guest profile database in use, and integration with existing systems is the top barrier to scaling AI, cited by 38%.

The infrastructure gaps compound. Ireckonu data from large hotel groups shows one in five guest profiles has no room reservation at all — restaurant, spa and golf guests who never appear in the PMS — leaving up to 20% of recognition and re-engagement opportunities untouched. Many PMS platforms create a new profile for every booking, cloud PMS migrations frequently carry only partial stay history, and OTA-masked email addresses make repeat guests strangers to hotel systems.

A further exposure: 60% of daily AI use in hotel chains runs through external tools such as ChatGPT, 27% through vendor-embedded systems and only 12% through internally built ones. Guest knowledge processed in browser tabs and vendor logs never returns to the hotel's own record.

What the fix looks like

Wyndham offers the most convincing production result of the season. Its voice agent, built on Salesforce Agentforce, now answers calls at properties where roughly a quarter of calls previously went unanswered. CEO Geoff Ballotti was explicit about the prerequisite: a half-billion-dollar technology-stack migration that began in 2016, consolidating dozens of PMS platforms down to two, before any AI was layered on top. About 85% of callers stay with the AI agent; the roughly 15% that transfer to a human are not failures — Wyndham reports a 15% lift in booking conversion and a 16% lift in ADR on those transferred calls.

The operational prescription that follows from both studies: resolve guest identity first, make every AI conversation write structured outputs back to the profile, build one governed guest memory layer with minimum-necessary access per agent, standardize the AI-to-human handoff packet, and disclose AI in the first message. Two dashboard metrics anchor it: the Repeat-Yourself Rate — the share of conversations where the guest re-supplies information the hotel already holds — and Manual Touches Per Stay, the staff-side mirror image.

Standards work is starting to lower the cost. Avalora's VAIA this week became the first AI assistant validated against the HTNG Express PMS specification, which standardizes access to a core of guest, reservation and room data. And from 1 November 2026, Singapore hotels can buy AI-enabled Digital Concierge solutions as a pre-approved category under Enterprise Singapore's EDGE for Productivity scheme.

Regulation tightens the deadline

Article 50 of the EU AI Act has applied since 2 August 2026, requiring chatbots and AI agents to tell people they are interacting with AI, at the latest at first interaction. The rule reaches non-EU businesses whose chatbots serve EU users — any APAC resort with European guests is in scope. In the Philippines, the Data Privacy Act (RA 10173) already requires lawful basis, proportionality and transparency for processing personal data. Against that backdrop, the 22% AI-disclosure figure is a compliance gap as much as a trust gap.

Hyatt's SVP of Data and AI, Pat Nestor, framed the industry's shift as moving from adoption to absorption, noting that hotel companies know what AI costs far better than what it earns. Agoda CTO Idan Zalzberg made the engineering case: "The real cost is not simply the model call" — it is the work that makes AI trustworthy in production. With 55% of chains planning guest-facing agents while personalization use declines, operators that fix identity resolution and memory write-back now will hold a measurable cost advantage when that wave arrives.