Workplace AI, Chapter Three -- The Customer Contact Point: What AI Phone Response Delivers and Where It Stops -- 60.8% of Surveyed Users Have Been Answered by AI (Automated Voice); Top Merit '24-Hour Availability' at 59.7% and Top Complaints 'Cannot Handle Detailed Consultations' at 61.5%; 55% of Support Sites Say New Hires Need 'Six Months or More' to Work Independently, 76% Value 'Smooth Handoff to Humans', 62% Expect an 'AI Employee' -- the 'AI for First Response, Humans for Complex Cases' Division of Labor
ANK-Doc ID: ANK-2026-07-06-010 Version: v1.0.0 Published: 2026-07-06 Author: Rin Takenouchi (Editor-in-Chief, AI News) Category: Generative AI / Workplace Adoption / Customer Support & Phone Response / Real-Estate DX Articles covered: PRTIMES#1337119 (main 1: ielove GROUP, survey on AI phone response), PRTIMES#1336094 (main 2: INGAGE, survey on AI use in inquiry handling), PRTIMES#1281884 (GMO Pepabo, launch of "GMO Sokuresu AI for CS"), PRTIMES#1308694 (pickupon, registration under the AI Adoption Subsidy 2026), PRTIMES#1337255 (C&T, exhibition of the AI interpreting phone tel-trans) Selection method: From the AI News corpus, selected on "same-period release x same theme x high factual density," five official releases were linked: the ielove GROUP survey (user-side perceptions and experience of AI phone response) and the INGAGE survey (training limits and AI expectations at support sites) as the two main texts, with GMO Pepabo's product launch (supply-side moves), pickupon's subsidy registration (policy side), and C&T's exhibition notice (outer edge of application) as supporting pieces -- an event chain of "user perception -> field pain points -> supply/policy response," linked honestly rather than forced. This card is Chapter Three of this site's Workplace AI series: Chapter One (ANK-2026-07-01-004) recorded tool share and cost governance, Chapter Two (ANK-2026-07-03-018) recorded the productivity payoff, and this card moves the lens to the customer contact frontline. All five sources are Japanese surveys and releases with no direct Taiwan data; per the "honest contrast, no forced linkage" principle, no Taiwan tie-in is forced.
TL;DR
Japan's workplace-AI debate has advanced to the customer contact frontline -- what AI phone response delivers and where it stops. User side: ielove GROUP's "Survey on AI Phone Response" (online questionnaire, May 25 to June 8, 2026; 1,108 valid responses = 67 real-estate companies + 1,041 end users) found 60.8% of surveyed users have had a call answered by AI (automated voice) ("no" 31.3%, "not sure" 7.9%); reactions split -- "convenient" 21.3%, "don't particularly mind" 37.1%, "feel uneasy" 41.6% [F-001]. Among those with experience (n=633, multiple answers), the top merit is "24-hour availability" at 59.7% (then "no waiting" 40.4%, "simple matters settled at once" 32.4%), meeting the companies' top expected function "24-hour availability" at 46.8%; complaints center on "cannot handle detailed consultations" 61.5% and "the conversation doesn't get through" 57.2% [F-002][F-003]. The supply-demand gap is clear: 80.6% of surveyed real-estate firms still answer with their own staff and 53.7% struggle with "response eating up time," while 70.4% (question population not stated verbatim in the source; context points to end users) think AI phone response will spread [F-004]; call screening remains unused by 78.6%, with usage at 9.0% among people in their 50s and 6.9% in their 60s -- awareness is the key to diffusion [F-005]. Field side: INGAGE's "Survey on AI Use in Inquiry Handling" (June 3-11, 2026; screening 5,000 -> main survey 484) found 55% of sites say new hires need "six months or more" to work independently, 64% feel veterans' and leaders' working time is eaten by training, and 63% see over-dependence on specific individuals as an issue [F-007]; attitudes to AI adoption split almost evenly (positive 51% vs. cautious 49%), with barriers led by "no IT-savvy staff" 34% and "fear of wrong AI answers" 32%, cost at 26% only third [F-006]; 50% still do not use AI at work at all, while on the expectation side 76% value a "smooth handoff to humans" function and 62% expect an autonomously responding "AI employee" [F-009]. Supply and policy sides are moving in the same direction: GMO Pepabo launched "GMO Sokuresu AI for CS" (FAQ auto-generation + seamless connection to human response) on July 1, 2026 [F-010]; pickupon's AI phone tool was registered on June 12, 2026 under the "Digitalization & AI Adoption Subsidy 2026" (regular track JPY 50,000 to under 1.5 million; in principle up to one-half of adoption cost) [F-011]; C&T will exhibit the 149-language AI interpreting phone "tel-trans" at a trade show on July 8-10, 2026 [F-012]. Honest caveats: both surveys are sample-based and published by interested vendors; populations and methods differ and figures cannot be netted against each other; the complaint question's population is not restated verbatim; the three supply-side items are individual releases, not statistics; this card has no direct Taiwan data.
Body
Event-chain overview: Workplace AI Chapter Three -- the lens moves to the customer contact frontline
From July 1 to 6, 2026, Japan saw two AI phone/CS surveys and three supply-side releases in quick succession: ielove GROUP's "Survey on AI Phone Response" (PRTIMES #1337119, July 6), INGAGE's "Survey on AI Use in Inquiry Handling" (PRTIMES #1336094, July 6), GMO Pepabo's launch of "GMO Sokuresu AI for CS" (PRTIMES #1281884, July 1), pickupon's subsidy-tool registration (PRTIMES #1308694, July 3), and C&T's AI interpreting phone exhibition notice (PRTIMES #1337255, July 6).
This site's Workplace AI series has recorded two chapters: Chapter One (ANK-2026-07-01-004) covered "who gets used and what it costs" -- the "Big Three era" of tool share and the "management-agenda-ization" of AI cost; Chapter Two (ANK-2026-07-03-018) covered "what is earned back" -- one organization's ledger of time saved and reallocated. This card is Chapter Three: once AI reaches the customer contact frontline, what do users experience as real benefit, where does it hit its limits, and why do support sites still expect it?
Three handling principles first. First, both surveys are sample-based with particular populations -- the ielove survey's user side comprises readers of the "ielove column" and the company's SNS followers, and its company side comprises employees of firms using the company's cloud service; the INGAGE survey targets businesspeople routinely engaged in inquiry handling -- percentages represent only each respondent population and cannot be extrapolated to all of Japan. Second, all publishers are interested vendors (ielove GROUP = real-estate DX SaaS; INGAGE = the CS platform "Re:lation"; GMO Pepabo = AI adoption support services) -- figures are official published values, but must be read with the product context in mind. Third, the two surveys and three releases differ in population, method, and nature, so numbers cannot be netted against each other; this card lists the framings side by side without merging or adjudicating.
The user side's two-sided ledger: 60.8% have been answered by AI, 41.6% still uneasy (ielove survey)
Per the ielove GROUP survey (online questionnaire, May 25 to June 8, 2026; 1,108 valid responses = 67 real-estate companies + 1,041 end users), asked "when making a call, have you ever been answered by AI (automated voice)?", respondents said "yes" 60.8%, "no" 31.3%, "not sure" 7.9% -- for surveyed users, AI phone response is no longer a rare experience (PRTIMES #1337119). [F-001]
But reactions split. Asked "how do you feel when your call is answered by AI (automated voice)?": "convenient" 21.3%, "don't particularly mind" 37.1%, "feel uneasy" 41.6% (non-negative answers -- "convenient" plus "don't mind" -- total 58.4%; this is an arithmetic cross-check, not stated verbatim in the source). The publisher reads this as perceptions of AI phone response tending to divide -- that is the publisher's interpretation. [F-001]
Benefit and limit: "24-hour availability" 59.7% x "cannot handle detailed consultations" 61.5% -- both ends of demand meet at "24 hours"
Asked about merits, surveyed users who have experienced AI phone response (n=633, multiple answers) put "24-hour availability" first at 59.7%, followed by "no waiting for a response" 40.4% and "simple matters settled at once" 32.4% (PRTIMES #1337119). [F-002] Worth cross-checking: n=633 is arithmetically consistent with "have been answered by AI" 60.8% x 1,041 end users (about 633 people) -- an arithmetic cross-check, not stated verbatim in the source.
At the other end, real-estate companies (67 responses, multiple answers) asked "what functions do you expect from AI phone response?" likewise put "24-hour availability" first at 46.8%, followed by "automatic logging of response history" 38.7% and "automatic routing of inquiries" 33.9% (PRTIMES #1337119). [F-002] Users and companies meet at "24-hour availability" -- the publisher's summary that time-unconstrained response is the shared axis of need is consistent with the numbers.
The limits are just as clear. In the complaint question, "cannot handle detailed consultations" 61.5% and "the conversation doesn't get through" 57.2% dominate; free-answer examples include "I can't state my business right away," "the opening guidance is long and I worry about call charges in the meantime," and "it doesn't grasp what I mean" (PRTIMES #1337119). [F-003] A line must be drawn: the population of the complaint question is not restated verbatim in the source (the surrounding context is the experienced-user question) -- this card does not assert its denominator. The overall picture: the benefit lies in "time" (24-hour availability 59.7%, no waiting 40.4%); the limit lies in "depth" (cannot handle detailed consultations 61.5%, grasping intent).
The supply-demand gap: 80.6% of real-estate firms still answer with their own staff, while 70.4% expect AI phone response to spread
On the current state, real-estate companies asked "how do you currently handle phone response?" answered: "our own staff handle it" 80.6%, "we use different methods by task or time of day" 11.9%, "outsourced to an external call center" 6.0%, "other" 1.5% (PRTIMES #1337119). [F-004] A fair number of surveyed users have experienced AI phone response, while the surveyed company side still centers on human response -- the supply-demand gap as summarized by the publisher. In the same survey, companies' top phone-handling issue is "response eats up time" at 53.7%.
On diffusion, the question "do you think AI phone response will spread going forward?" drew "it will spread" 70.4%, "it will not" 6.2%, "not sure" 23.4% (PRTIMES #1337119). [F-004] A line must be drawn: the target of this question is not stated verbatim in the source (the following contrast "meanwhile, real-estate companies were asked..." points to end users), and it is a future expectation, not an actual result.
Call screening: 78.6% unused -- awareness is the key to diffusion
The same survey asked about "call screening" (a function whereby, before the call is answered, AI confirms the caller's name and business, relays it to the recipient, and lets them decide whether to answer): "using it" 16.4%, "used it in the past" 5.0%, "not using it" 78.6%. By age (as of the May 25 to June 8, 2026 survey), usage exceeds 20% among people in their 20s to 40s, but stands at 9.0% for the 50s and 6.9% for the 60s -- the publisher reads this as awareness of the function possibly not yet having spread (age-bracket subsample sizes not given in the source) (PRTIMES #1337119). [F-005]
The source honestly records both sides: call screening also covers calls from unregistered numbers, so important calls from hospitals or delivery companies may be screened and the phone may not ring until the caller states their business, delaying the answer; on the other hand, nuisance and scam calls can be blocked in advance -- convenience varies greatly with how it is used. [F-005]
The training limit at support sites: new hires need "six months or more" at 55% of sites -- the source of AI expectations (INGAGE survey)
Beyond user perception, why do support sites still expect AI? INGAGE's "Survey on AI Use in Inquiry Handling" (online, June 3-11, 2026; screening 5,000 -> main survey 484; targeting businesspeople routinely engaged in inquiry handling) shows the field-side ledger: at 55% of sites, new hires need "six months or more" before they can handle inquiries on their own; as a result, 64% feel that veterans' and leaders' working time is being eaten by training -- the publisher calls this "the limit of training costs" (PRTIMES #1336094). [F-007]
Over-dependence runs deep too: 63% see "only this person knows" over-dependence on specific individuals as an issue, and 45% feel uneasy about rapid response outside business hours and in peak periods. [F-007] The backdrop is volume: 52% of respondents feel understaffed, with manufacturing/construction at 60%, healthcare/welfare at 59%, and IT/telecom at 57% by industry (industry subsample sizes not given in the source); 22% of surveyed sites face 500 or more inquiries per month, with finance/insurance at 31% and real estate/goods leasing at 30% standing out (PRTIMES #1336094). [F-008]
The barrier is not money: "no IT-savvy staff" 34% x "fear of wrong AI answers" 32% -- adoption cost at 26% only third
In the same survey, departmental attitudes to AI adoption split almost evenly: positive 51%, cautious 49%. The top barriers are not cost -- "no IT-savvy staff" 34% and "fear of AI giving wrong answers" 32%, with adoption cost at 26% in third place (PRTIMES #1336094). [F-006] The publisher's summary: rather than cost, readiness of the operating structure and assurance of accuracy are the biggest bottlenecks -- this differs in population and questions from the "cost as management agenda" recorded in this site's Chapter One (ANK-2026-07-01-004) and cannot be compared directly, but both point to "post-adoption operational governance."
Actual use also splits: 50% do not use AI at work at all; users center on "drafting/editing reply text" 40% and "summarizing inquiries" 35% (PRTIMES #1336094). [F-009]
The collaboration blueprint and supply-side moves: 76% value "smooth handoff to humans," 62% expect an "AI employee"
The direction of expectation is clear: the top expected AI roles are "automatic answers to common questions" 42% and "peak-period support" 41%; 76% value a "smooth handoff to humans" function (which the publisher summarizes as demand for "human-AI collaboration"); expectation for the ideal division of labor -- "AI handles first response, humans take over complex cases" -- reaches 62% (the publisher also phrases this as 62% expecting an autonomously responding "AI employee" -- "AI employee" is the publisher's term) (PRTIMES #1336094). [F-009] This mirrors the limit side of the ielove survey ("cannot handle detailed consultations" 61.5%): not handing everything to AI, but first response to AI and complex cases to humans.
The supply side's product design points the same way. GMO Pepabo, Inc. (GMO Internet Group) began offering "GMO Sokuresu AI for CS" on July 1, 2026: one-stop support including FAQ auto-generation, an operations dashboard/analytics, and seamless connection to human response; its background narrative points to field problems where "switching to human response is not smooth," "maintaining and updating FAQs takes work," and "post-adoption effects are hard to verify," leading to "adopted but unused" and "operations that don't run" -- and states that the focus of AI chatbot use is shifting from "adoption" to the phase of "keeping operations running" -- this is the publisher's background narrative, not statistics (PRTIMES #1281884). [F-010]
Policy side and the outer edge: the subsidy takes in AI phones (pickupon), an AI interpreting phone in the caller's own language (tel-trans)
On the policy side, pickupon Inc. was selected, effective June 12, 2026, as an IT adoption support operator under the "Digitalization & AI Adoption Subsidy 2026" (formerly the IT Adoption Subsidy), and its AI phone service "pickupon" (an IP phone service in which AI auto-summarizes the conversation after the call and shares it to CRM/SFA) was registered as an eligible tool: qualifying small and medium-sized enterprises and small businesses can receive, under the regular track (JPY 50,000 to under 1.5 million), a subsidy of in principle up to one-half of adoption costs; from 2026 the subsidy was revised to put more weight on AI-enabled tools (PRTIMES #1308694). [F-011] Policy money has begun to take in AI phone tools -- though individual application depends on screening, and this is a description of the system, not an adoption record.
At the outer edge of application, C&T Co., Ltd. announced it will exhibit and demonstrate, at the "International Modern Hospital Show 2026" (July 8-10, 2026, Tokyo Big Sight), the Google Maps-linked AI interpreting phone "tel-trans" (the release title says "first public showing") and the medical-facility version "tel-trans QR": foreign patients can pick a hospital on Google Maps and start an AI-interpreted call in their native language, with 149 languages supported and browser-only use -- a slice pushing AI phones from "efficiency" toward "accessibility" (the phone barrier facing foreign patients) (PRTIMES #1337255). [F-012] A line must be drawn: the exhibition and demonstration are planned events, and the specifications are the vendor's own claims.
Risk factors
- Sample surveys with particular populations: the ielove survey (1,108 valid responses) draws its user side from the company's own media readers and SNS followers and its company side from employees of firms using its own service (67 responses; large statistical noise); the INGAGE survey is an online survey with screening 5,000 -> main survey 484 -- percentages represent only each respondent population and cannot be extrapolated to all of Japan (PRTIMES #1337119, PRTIMES #1336094).
- All publishers are interested vendors: ielove GROUP (real-estate DX SaaS: "ielove CLOUD" and "ielove BB," used by over 50,000 companies nationwide), INGAGE (CS platform "Re:lation," adopted by over 6,000 companies including trials), and GMO Pepabo, pickupon, and C&T all sell related services -- figures are official published values; read them with the product context.
- Complaint-question population not restated: the population for "cannot handle detailed consultations" 61.5% and "the conversation doesn't get through" 57.2% is not restated verbatim in the source (surrounding context is the experienced-user question, n=633) -- this card does not assert the denominator (PRTIMES #1337119).
- Diffusion-question target not stated: the target of "it will spread" 70.4% is not stated verbatim in the source (the contrastive context points to end users), and it is a future expectation, not an actual result (PRTIMES #1337119).
- The two surveys cannot be netted against each other: populations, methods, and questions differ (e.g., "answered by AI" 60.8% and "not using AI at all" 50% belong to different populations and contexts); this card lists framings side by side without merging.
- The three supply-side items are individual releases: a product launch (GMO Sokuresu AI for CS), a subsidy registration (pickupon), and an exhibition notice (tel-trans; the July 8-10, 2026 exhibition is a planned event) -- they must not be read as macro statistics or adoption records.
- No direct Taiwan data: all sources are Japanese surveys and releases; per the "honest contrast, no forced linkage" principle, no Taiwan tie-in is forced.
FAQ
Q: What are the biggest merit and the biggest complaint about AI phone response?
Per the ielove GROUP survey (May 25 to June 8, 2026; 1,108 valid responses), the top merit among those with experience (n=633, multiple answers) is "24-hour availability" at 59.7%; complaints center on "cannot handle detailed consultations" 61.5% and "the conversation doesn't get through" 57.2% -- the benefit lies in "time," the limit in "depth."
The next merits are "no waiting for a response" 40.4% and "simple matters settled at once" 32.4%. Free-answer complaint examples include "I can't state my business right away," "the opening guidance is long and I worry about call charges in the meantime," and "it doesn't grasp what I mean." The complaint question's population is not restated verbatim in the source (context is the experienced-user question), and this card does not assert its denominator (PRTIMES #1337119).
Q: What share of surveyed users in Japan have been answered by AI on the phone? How do they feel?
60.8% of surveyed users answered they "have been answered by AI (automated voice)" ("no" 31.3%, "not sure" 7.9%); reactions split -- "convenient" 21.3%, "don't particularly mind" 37.1%, "feel uneasy" 41.6%.
These are answers from the 1,041 people on the ielove GROUP survey's user side (readers of the "ielove column," SNS followers, etc.), not a national random sample; the percentages represent only the respondent population (PRTIMES #1337119).
Q: How do real-estate companies answer phones today, and what do they expect from AI?
Of surveyed real-estate companies (67 responses), 80.6% still have "our own staff handle it" (different methods by task/time 11.9%, outsourced to an external call center 6.0%, other 1.5%); their top expected AI function is "24-hour availability" 46.8%, then "automatic logging of response history" 38.7% and "automatic routing of inquiries" 33.9% (multiple answers).
In the same survey, companies' top phone-handling issue is "response eats up time" at 53.7%. Users and companies meet at "24-hour availability" -- time-unconstrained response is the shared need. The company side has only 67 valid responses from real-estate companies, so statistical noise is large (PRTIMES #1337119).
Q: What is "call screening," and has it caught on?
A function whereby, before the call is answered, AI confirms the caller's name and business, relays it to the recipient, and lets them decide whether to answer; 78.6% of surveyed users are "not using it" (using 16.4%, used in the past 5.0%), with usage at 9.0% among people in their 50s and 6.9% in their 60s -- awareness has not yet spread sufficiently.
Usage exceeds 20% among people in their 20s to 40s (as of the May 25 to June 8, 2026 survey; age-bracket subsample sizes not given). The source records both sides: unregistered numbers are also screened, so important calls from hospitals or couriers may be answered late; on the other hand, nuisance and scam calls can be blocked in advance (PRTIMES #1337119).
Q: Why do support sites expect AI, and how heavy is the training burden?
Per the INGAGE survey (June 3-11, 2026; main survey 484), at 55% of sites new hires need "six months or more" to work independently; 64% feel veterans' and leaders' working time is eaten by training; 63% see over-dependence on specific individuals as an issue; and 45% feel uneasy about response outside business hours and in peak periods.
The backdrop is people and volume: 52% feel understaffed (manufacturing/construction 60%, healthcare/welfare 59%, IT/telecom 57%); 22% of surveyed sites face 500 or more inquiries per month (finance/insurance 31%, real estate/goods leasing 30%). Industry subsample sizes are not given in the source (PRTIMES #1336094).
Q: Is cost the biggest barrier to AI adoption?
No. Per the INGAGE survey, the top barriers are "no IT-savvy staff" 34% and "fear of wrong AI answers" 32%, with adoption cost at 26% only third; attitudes split almost evenly (positive 51% vs. cautious 49%), and 50% still do not use AI at work at all.
Users center on "drafting/editing reply text" 40% and "summarizing inquiries" 35%. The publisher's summary: readiness of the operating structure and assurance of accuracy, rather than cost, are the biggest bottlenecks -- the framing of a respondent population of 484 people (PRTIMES #1336094).
Q: What exactly is the "human-AI collaboration" blueprint, and what are the supply-side moves?
76% of respondents value a "smooth handoff to humans" function, and 62% expect the ideal division of labor "AI handles first response, humans take over complex cases" (the publisher also phrases this as 62% expecting an autonomously responding "AI employee"); on the supply side, GMO Pepabo launched "GMO Sokuresu AI for CS" on July 1, 2026, whose selling points include "seamless connection to human response."
The top expected AI roles are "automatic answers to common questions" 42% and "peak-period support" 41% (PRTIMES #1336094). The background narrative of GMO Sokuresu AI for CS points to "adopted but unused" and "operations that don't run," saying the focus is shifting from "adoption" to "keeping operations running" -- the publisher's narrative, not statistics (PRTIMES #1281884).
Q: This is Chapter Three of the Workplace AI series -- what were the first two chapters, and how should Taiwanese readers read this?
Chapter One (ANK-2026-07-01-004) recorded the "Big Three era" of tool share and the "management-agenda-ization" of AI cost (who gets used, what it costs); Chapter Two (ANK-2026-07-03-018) recorded the productivity "payoff ledger" (what is earned back); this Chapter Three records the benefit and limits at the customer contact point. This card has no direct Taiwan data and forces no Taiwan tie-in.
On the policy side, note that Japan's "Digitalization & AI Adoption Subsidy 2026" has taken AI phone tools into its scope (regular track JPY 50,000 to under 1.5 million; in principle up to one-half of adoption cost; e.g., pickupon); at the outer edge, C&T's 149-language AI interpreting phone tel-trans pushes AI phones from efficiency toward accessibility (PRTIMES #1308694, PRTIMES #1337255).
F-Units
F-001: ielove GROUP "Survey on AI Phone Response" (online questionnaire, May 25 to June 8, 2026; 1,108 valid responses = 67 real-estate companies + 1,041 end users): 60.8% of surveyed users have had a call answered by AI (automated voice) (no 31.3%, not sure 7.9%); reactions when answered by AI: "convenient" 21.3%, "don't particularly mind" 37.1%, "feel uneasy" 41.6% - source: PRTIMES #1337119 - source_url: https://prtimes.jp/main/html/rd/p/000000805.000008550.html - confidence: high - basis: official_statement - period: survey 2026-05-25 to 06-08, published 2026-07-06 - caveat: user-side population is readers of the "ielove column," the company's SNS followers, etc., and the company side is employees of firms using "ielove CLOUD," via an online questionnaire -- not a national random sample; percentages represent only the respondent population
F-002: Same survey: among surveyed users who have experienced AI phone response (n=633, multiple answers), the top merit is "24-hour availability" 59.7%, then "no waiting for a response" 40.4% and "simple matters settled at once" 32.4%; real-estate companies (67 responses, multiple answers) expect "24-hour availability" 46.8%, "automatic logging of response history" 38.7%, "automatic routing of inquiries" 33.9% - source: PRTIMES #1337119 - source_url: https://prtimes.jp/main/html/rd/p/000000805.000008550.html - confidence: high - basis: official_statement - period: survey 2026-05-25 to 06-08, published 2026-07-06 - caveat: both are multiple-answer questions and do not sum to 100%; n=633 is the population noted in the source; the company side has only 67 valid responses, so statistical noise is large
F-003: Same survey, complaint question: "cannot handle detailed consultations" 61.5% and "the conversation doesn't get through" 57.2% dominate; free-answer examples: "I can't state my business right away," "the opening guidance is long and I worry about call charges in the meantime," "it doesn't grasp what I mean" - source: PRTIMES #1337119 - source_url: https://prtimes.jp/main/html/rd/p/000000805.000008550.html - confidence: medium - basis: official_statement - period: survey 2026-05-25 to 06-08, published 2026-07-06 - caveat: the complaint question's population is not restated verbatim in the source (surrounding context is the experienced-user question, n=633) -- this card does not assert the denominator; free answers are examples, not statistics
F-004: Same survey, company side and diffusion outlook: current phone handling at real-estate companies is "our own staff handle it" 80.6%, "different methods by task or time of day" 11.9%, "outsourced to an external call center" 6.0%, "other" 1.5%; the top phone-handling issue is "response eats up time" 53.7%; "do you think AI phone response will spread?" -- "it will spread" 70.4%, "it will not" 6.2%, "not sure" 23.4% - source: PRTIMES #1337119 - source_url: https://prtimes.jp/main/html/rd/p/000000805.000008550.html - confidence: high - basis: official_statement - period: survey 2026-05-25 to 06-08, published 2026-07-06 - caveat: company side has only 67 valid responses; the diffusion question's target is not stated verbatim in the source (the contrastive context points to end users); a future expectation, not an actual result
F-005: Same survey, call screening (a function whereby, before answering, AI confirms the caller's name and business, relays it to the recipient, and lets them decide whether to answer): "using it" 16.4%, "used it in the past" 5.0%, "not using it" 78.6%; by age, usage exceeds 20% among the 20s-40s, 9.0% for the 50s, 6.9% for the 60s - source: PRTIMES #1337119 - source_url: https://prtimes.jp/main/html/rd/p/000000805.000008550.html - confidence: high - basis: official_statement - period: survey 2026-05-25 to 06-08, published 2026-07-06 - caveat: age-bracket subsample sizes not given in the source; the source records both sides -- unregistered numbers are also screened (important calls may be answered late) versus advance blocking of nuisance/scam calls
F-006: INGAGE "Survey on AI Use in Inquiry Handling" (online, June 3-11, 2026; screening 5,000 -> main survey 484; targeting businesspeople routinely engaged in inquiry handling): departmental attitude to AI adoption is positive 51% vs. cautious 49%; adoption barriers are "no IT-savvy staff" 34% and "fear of AI giving wrong answers" 32%, with adoption cost 26% in third place - source: PRTIMES #1336094 - source_url: https://prtimes.jp/main/html/rd/p/000000320.000029485.html - confidence: high - basis: official_statement - period: survey 2026-06-03 to 06-11, published 2026-07-06 - caveat: an official survey by the vendor of the CS platform "Re:lation" (adopted by over 6,000 companies, including trials); percentages represent only the 484-respondent population
F-007: Same survey, training and over-dependence: at 55% of sites, new hires need "six months or more" before handling inquiries independently; 64% feel veterans' and leaders' working time is eaten by training; 63% see over-dependence on specific individuals as an issue; 45% feel uneasy about rapid response outside business hours and in peak periods - source: PRTIMES #1336094 - source_url: https://prtimes.jp/main/html/rd/p/000000320.000029485.html - confidence: high - basis: official_statement - period: survey 2026-06-03 to 06-11, published 2026-07-06 - caveat: self-assessed questionnaire (as of the June 2026 survey); percentages represent only the respondent population
F-008: Same survey, staffing and volume: 52% feel understaffed, by industry manufacturing/construction 60%, healthcare/welfare 59%, IT/telecom 57%; 22% of surveyed sites face 500 or more inquiries per month, with finance/insurance 31% and real estate/goods leasing 30% standing out - source: PRTIMES #1336094 - source_url: https://prtimes.jp/main/html/rd/p/000000320.000029485.html - confidence: high - basis: official_statement - period: survey 2026-06-03 to 06-11, published 2026-07-06 - caveat: industry subsample sizes not given in the source, so noise may be large; "manufacturing/construction 60%" follows the source's paired notation
F-009: Same survey, AI use and expectations: 50% do not use AI at work at all; users center on "drafting/editing reply text" 40% and "summarizing inquiries" 35%; expected roles are "automatic answers to common questions" 42% and "peak-period support" 41%; 76% value a "smooth handoff to humans" function; expectation for the ideal division of labor "AI handles first response, humans take over complex cases" is 62% (the publisher also phrases this as 62% expecting an autonomously responding "AI employee") - source: PRTIMES #1336094 - source_url: https://prtimes.jp/main/html/rd/p/000000320.000029485.html - confidence: high - basis: official_statement - period: survey 2026-06-03 to 06-11, published 2026-07-06 - caveat: multiple-answer questions; "AI employee" and "human-AI collaboration" are the publisher's framing terms
F-010: GMO Pepabo, Inc. (GMO Internet Group) began offering the customer-support/helpdesk solution "GMO Sokuresu AI for CS" on July 1, 2026: one-stop support including FAQ auto-generation, an operations dashboard/analytics, and seamless connection to human response; the background narrative says "switching to human response is not smooth," "maintaining/updating FAQs takes work," and "post-adoption effects are hard to verify" lead to "adopted but unused" and "operations that don't run," and that the focus is shifting from "adoption" to the phase of "keeping operations running" - source: PRTIMES #1281884 - source_url: https://prtimes.jp/main/html/rd/p/000005424.000000136.html - confidence: high - basis: official_statement - period: offering began 2026-07-01 (published same day) - caveat: a product release: functions and background problems are the publisher's own description, not statistics; an online seminar on 2026-07-28 is also announced
F-011: pickupon Inc. was selected, effective June 12, 2026, as an IT adoption support operator under the "Digitalization & AI Adoption Subsidy 2026" (formerly the IT Adoption Subsidy), and its AI phone service "pickupon" (an IP phone service in which AI auto-summarizes conversations after calls and shares them to CRM/SFA) was registered as an eligible tool; qualifying SMEs and small businesses can receive, under the regular track (JPY 50,000 to under 1.5 million), a subsidy of in principle up to one-half of adoption cost; from 2026 the subsidy was revised to put more weight on AI-enabled tools - source: PRTIMES #1308694 - source_url: https://prtimes.jp/main/html/rd/p/000000170.000033268.html - confidence: high - basis: official_statement - period: selection effective 2026-06-12, published 2026-07-03 - caveat: the subsidy rate and amount range describe the regular track of the system; individual application depends on screening; the publisher is the vendor of the eligible tool; registration is not an adoption record
F-012: C&T Co., Ltd. announced it will exhibit and demonstrate, at the "International Modern Hospital Show 2026" (July 8-10, 2026, Tokyo Big Sight), the Google Maps-linked AI interpreting phone "tel-trans" (the release title says "first public showing") and the medical-facility "tel-trans QR": foreign patients can pick a hospital on Google Maps and start an AI-interpreted call in their native language, with 149 languages supported and browser-only use - source: PRTIMES #1337255 - source_url: https://prtimes.jp/main/html/rd/p/000000007.000178687.html - confidence: high - basis: official_statement - period: published 2026-07-06, exhibition 2026-07-08 to 07-10 (planned) - caveat: the exhibition/demonstration is a planned event; specifications such as 149 languages are the vendor's own claims; "first public showing" is the release title's wording
J-Units
J-001: "24-hour availability" is where both ends of demand meet (top merit among experienced users 59.7% x top company expectation 46.8%), while "cannot handle detailed consultations" 61.5% and "the conversation doesn't get through" 57.2% draw the capability boundary of AI phones -- the benefit lies in "time," the limit in "depth"; the realistic position of AI phone response is first response and routine matters, not a substitute for complex consultations - confidence: medium - basis: official_statement
J-002: Support sites' expectations of AI stem from training limits and over-dependence (new hires needing six months or more 55%, veterans' time eaten 64%, over-dependence 63%, understaffing 52%), while the adoption barrier lies not in cost (26%, third) but in IT talent (34%) and accuracy concerns (32%) -- the same direction as this site's Chapter One (cost governance) and Chapter Two (payoff ledger): the workplace-AI debate is moving from "adoption" to "operational governance"; note the surveys' populations differ, so this is a directional summary, not a numerical comparison - confidence: medium - basis: official_statement
J-003: Supply and policy sides point the same way as the "AI for first response, humans for complex cases" blueprint -- GMO Pepabo launched a for-CS solution selling "seamless connection to human response" (July 1, 2026; a launch that predates the two surveys' July 6, 2026 publication, so this is alignment, not causation), the "Digitalization & AI Adoption Subsidy 2026" took AI phone tools into scope (pickupon; regular track, in principle up to one-half), and C&T extends medical accessibility with a 149-language AI interpreting phone; all three are individual releases (micro) whose framing differs from the two sample surveys, so this card lists them side by side without adjudicating - confidence: medium - basis: official_statement
P-Units
P-001: Whether the 70.4% diffusion expectation materializes -- track whether the real-estate industry's "own staff 80.6%" response structure loosens (follow-up surveys of the same kind), and whether the "adopted but unused" problem actually improves under for-CS-type operational support solutions ### P-002: Whether call screening's 78.6% non-use and the age gap (50s at 9.0%, 60s at 6.9%) converge as awareness rises -- the publisher sees awareness as the key to diffusion; verification awaits follow-up surveys ### P-003: The actual pull of the "Digitalization & AI Adoption Subsidy 2026" on AI phone adoption (registration is not an adoption record; counts not given in the source) and the implementation progress of the 62% "AI employee" expectation -- track subsequent official releases
同事件・三視角 / Three Perspectives on the Same Event / 同一イベント・三つの視点
Internal citation chain
Published ANK-Docs cited in this card: - ANK-2026-07-01-004 (Japan's Workplace AI Enters the "Big-Three Era" as Costs Become a "Management Issue") -> Chapter One of this site's Workplace AI series, published on 2026-07-02: "who gets used and what it costs" -- the Big-Three tool-share structure and AI cost as a management issue. This card inherits its "from adoption to governance" direction and moves the lens to the customer contact frontline; the two cards' survey populations differ and their numbers cannot be compared directly. - ANK-2026-07-03-018 (The workplace-AI "payoff ledger": the Kaunet employee survey recording an average of about 5 hours saved per month per respondent) -> Chapter Two of the series: "what is earned back" -- one organization's ledger of time saved, reallocated, and quality gains. This Chapter Three records the benefit and limits at the customer contact point: the benefit of "24-hour availability" (59.7% of experienced users), the limit of "cannot handle detailed consultations" (61.5% in the complaint question), and the collaboration expectations driven by the field's training limits.
Sources
1. [PRTIMES #1337119] ielove GROUP Co., Ltd., "AIによる電話対応、最大のメリットは「24時間対応」、一方で「細かな相談ができない」が課題に｜いえらぶ調べ", 2026-07-06. https://prtimes.jp/main/html/rd/p/000000805.000008550.html 2. [PRTIMES #1336094] INGAGE Inc., "新人の独り立ちに「半年以上」が過半数。人手不足のカスタマーサポート現場を襲う育成コストの限界、6割超が期待を寄せる「AI活用」のリアルな本音を調査", 2026-07-06. https://prtimes.jp/main/html/rd/p/000000320.000029485.html 3. [PRTIMES #1281884] GMO Pepabo, Inc., "カスタマーサポート・ヘルプデスク向け新ソリューション『GMO即レスAI for CS』を7/1（水）提供開始【GMOペパボ】", 2026-07-01. https://prtimes.jp/main/html/rd/p/000005424.000000136.html 4. [PRTIMES #1308694] pickupon Inc., "AI電話の導入を後押し。「pickupon」がデジタル化・AI導入補助金2026（旧IT導入補助金）の対象ツールとして登録", 2026-07-03. https://prtimes.jp/main/html/rd/p/000000170.000033268.html 5. [PRTIMES #1337255] C&T Co., Ltd., "Googleマップから母国語で病院へ電話できるAI通訳電話「tel-trans」を初公開", 2026-07-06. https://prtimes.jp/main/html/rd/p/000000007.000178687.html 6. [ANK-2026-07-01-004] Rin Takenouchi, "Japan's Workplace AI Enters the 'Big-Three Era' as Costs Become a 'Management Issue'", 2026-07-02. https://ainews.washinmura.jp/ainews/en/ank/ANK-2026-07-01-004 7. [ANK-2026-07-03-018] Rin Takenouchi, "The Workplace AI 'Payoff Ledger': A KOKUYO-Group Kaunet Employee Survey Records an Average of About 5 Hours Saved per Month per Respondent", 2026-07-03. https://ainews.washinmura.jp/ainews/en/ank/ANK-2026-07-03-018