AI, EQ & Communication News — 2026-10-20 (Tue) Evening News

Overview

Headlines from the past 48 hours cover emotion-reading AI, voice-based conversation practice, and the way people talk with and about chatbots. Several items concern users setting boundaries with ChatGPT or noticing its phrasing, while new products offer spoken AI practice for English and high-stakes conversations. No tracked public figures appear in the material.

Emotion-reading AI

- A perspective piece proposes a testable framework for "super-emotional AI" (Bioengineer.org). - A webinar for local governments covers using AI speech recognition to protect staff from customer harassment ("kasuhara") and recent window-service digital transformation cases, and is available on demand (Niconico News). 【AI empathy and emotional intelligence】 No notable developments. 【Human EQ and skills in the AI era】 No notable developments. 【Changes in conversation and relationships】 - An item from thhsclassic.com describes how students at one school are setting boundaries with AI, under the headline "ChatGPT is not your friend" (thhsclassic.com). - A trend in which users share "five things I always say" to ChatGPT offers a view into people's chats (Mashable). - Forbes runs a piece titled "You Have 9 Minutes To Build Trust With AI. Will ChatGPT Recommend You?" (Forbes). - The AI English conversation app Speak has added a phone-call feature that remembers past conversations and lets users speak freely (d Menu News). - Ecommunication has launched "ECOM AI Lesson," which uses voice-based conversational AI (ICT Education News).

Words written and spoken by AI

- Commentator Matt Walsh accused actor Matthew McConaughey of using AI in a viral pre-election video, saying "This Is the Most ChatGPT Speech." This is Walsh's claim; the headline does not confirm that AI was used (NewsBreak).

Practice speaking and coaching

- Convrsy has emerged from stealth to help organizations test strategies and rehearse high-stakes conversations with AI (Yahoo Finance). - An article on AI job interviews advises candidates to stop expecting to be "read between the lines" and to speak on the premise that only the words they say will be conveyed (Yahoo! News Japan).

Research and surveys

No notable developments. 【Japan's EQ, communication and AI】 - Ecom has begun a voice-based conversational AI English lesson service (J-CAST News).

Statements by key figures

No notable developments. 【Integrated analysis: what is happening】 The headlines share one structure: conversation with AI is becoming a space that people practice in, set limits on, and scrutinize. Practice tools (Speak, ECOM AI Lesson, Convrsy) move rehearsal into spoken, voice-based AI, and the AI interview advice points the same way, treating spoken words as the only signal that gets evaluated. At the same time, the school boundary item, the chat-habit trend and the McConaughey accusation show people paying attention to how AI-influenced language sounds and how far to treat a chatbot as a companion. The Forbes headline suggests trust in AI is also being framed as something organizations must earn from the AI itself. This likely means that spoken fluency and trust signals are being shaped for AI audiences as well as human ones. The Walsh accusation is an unverified claim, so it shows suspicion of AI-sounding language rather than evidence of AI use.

What to watch

- Whether the proposed testable framework for "super-emotional AI" leads to concrete evaluation methods, which would show if emotion-reading claims can be checked (inference). - Whether voice-based practice tools, now launching in several forms, add features such as memory of past conversations, as Speak has done (inference). - Whether accusations that public speech is AI-written become more common as people attend to "ChatGPT-sounding" language (inference).