AI watch by topic
AI in customer service
Across Europe, artificial intelligence already handles 52 % of customer service contacts, according to a 2026 Twilio study, with companies expecting that figure to reach 60 % by 2027. Named deployments, from large insurers to small fundraising platforms, show measurable gains in call volumes and handling time, but companies and their customers report very different levels of satisfaction with the results, and some companies have had to reintroduce human staff after automation eroded trust.
AI already handles half of customer contacts in Europe
Across Europe, artificial intelligence already handles a large share of customer service contacts. According to a Twilio study covering France, Germany, Italy, Spain and the UK, AI managed 52 % of customer interactions in the EMEA region in 2026, a figure companies expect to rise to 60 % by 2027. European adoption, however, trails the rest of the world: the Freshworks Customer Service Benchmark Report 2026 found that only 40.8 % of European customer service teams had deployed automated tools, against 44 % globally. Where European teams had deployed them, though, 81 % handled more requests without adding staff, compared with 65.3 % worldwide.
-
Relation Client Mag · 13 October 2026 Customer service and AI: Twilio study points to a perception gap Read the article on the publisher’s website ↗ -
Comarketing-News · 12 October 2026 Customer service: 60% of requests will go through AI by 2027 Read the article on the publisher’s website ↗ -
Relation Client Mag · 14 October 2026 Customer service: AI boosts European productivity Read the article on the publisher’s website ↗
Our summaries: Relation Client Mag (1, 2), Comarketing-News
Companies and customers see AI’s reliability very differently
Companies and the customers they serve do not agree on how well AI is working. In the same Twilio study, 84 % of European decision-makers said their customers were satisfied with automated tools, but only 48 % of consumers shared that view; customers welcomed the time saved (81 %) but largely refused to let a robot initiate a loan (53 %). Infobip’s CX Maturity Report 2026 found that 76 % of European brands cite low user trust as the main barrier to deploying AI in customer experience, against 36 % citing technical integration problems. In France, a Hostinger-commissioned survey found 87 % of respondents still prefer a human adviser, rising to 99 % among those over 65.
-
Relation Client Mag · 13 October 2026 Customer service and AI: Twilio study points to a perception gap Read the article on the publisher’s website ↗ -
Comarketing-News · 12 October 2026 Customer service: 60% of requests will go through AI by 2027 Read the article on the publisher’s website ↗ -
HayCanal · 14 September 2026 User trust is a barrier to integrating AI into the customer experience Read the article on the publisher’s website ↗ -
Journal du Net · 17 September 2026 The AI Act and the Digital Omnibus: a postponement that exempts no one Read the article on the publisher’s website ↗
Our summaries: Relation Client Mag, Comarketing-News, HayCanal, Journal du Net
Named deployments report measurable time and call savings
Several companies report concrete figures. At AXA Health in the UK, the ALAN agent, built on Verint, supports over 1,000 contact-centre agents and handles around 55,000 calls a week, cutting average call-handling time by 60 seconds through automated summaries. At Zeffy, a Canadian fundraising platform, a first layer of AI, built on its own knowledge base, now handles about 80 % of customer support tickets. Joe AI, a French start-up serving estate agencies, says it processes several hundred thousand communications a month for over 500 French clients, aiming to cover up to 95 % of inbound requests. In the US, ASAPP’s agent rebooked more than 3,800 airline passengers in six days during a winter storm.
-
Brief IA · 11 September 2026 AXA rolls out its AI agents: why insurance is changing scale Read the article on the publisher’s website ↗
-
Maddyness · 28 September 2026 How Zeffy, the North American “cousin” of HelloAsso, is doubling in size every year thanks to AI Read the article on the publisher’s website ↗ -
Maddyness · 1 October 2026 Joe AI raises €2 million to automate customer relations in real estate Read the article on the publisher’s website ↗ -
Unite.AI · 15 September 2026 Katie Stein, CEO of ASAPP – Interview Series Read the article on the publisher’s website ↗
Smaller firms can access AI without heavy investment
Smaller companies are also adopting AI in customer service without heavy upfront investment. Vodafone Empresas says it has deployed AI agents to more than 100,000 SME clients in Spain, automating call handling, appointments and incident resolution through a pay-as-you-go subscription that avoids a heavy initial investment. This matches a broader pattern: a European Commission Eurobarometer found that 42 % of EU SMEs already use AI daily, mostly on a trial basis, and that customer service is their leading use case, at 47 %, ahead of logistics and workforce planning at 30 %. In Spain, 42 % of SMEs reported using AI for logistics specifically, above the EU average, while 24 % planned to adopt AI in future.
-
Economía Digital · 5 October 2026 Vodafone rolls out AI agents to more than 100,000 SMEs Read the article on the publisher’s website ↗ -
Cadena de Suministro · 15 October 2026 42% of Spanish SMEs use artificial intelligence in logistics Read the article on the publisher’s website ↗
Our summaries: Economía Digital, Cadena de Suministro
Even motivated adopters hit data and trust limits
Even companies committed to AI in customer service report limits. At Salesforce, chief executive Marc Benioff said AI already performs 30 % to 50 % of internal work, including customer service, during a period in which the company was linked to the loss of 4,000 support roles; Klarna, after cutting its human support team, reintroduced staff for some cases once the change began eroding customer trust. A Forrester Consulting study for Klaviyo, covering 163 French e-commerce and retail marketing managers, found that only 16 % had integrated agentic AI across all their customer engagement flows, while 23 % were still at pilot stage. The main obstacles cited were a lack of clean, unified data (31 %), insufficient internal skills (29 %), and compliance or integration problems (28 %).
-
iProfesional · 3 August 2026 The automation ceiling and its negative impact on user experience Read the article on the publisher’s website ↗ -
Relation Client Mag · 30 September 2026 Agentic AI: French companies held back by their CRM Read the article on the publisher’s website ↗
Our summaries: iProfesional, Relation Client Mag
Articles cited
- 42% of Spanish SMEs use artificial intelligence in logistics Cadena de Suministro, 15 October 2026
- Customer service: AI boosts European productivity Relation Client Mag, 14 October 2026
- Customer service and AI: Twilio study points to a perception gap Relation Client Mag, 13 October 2026
- Customer service: 60% of requests will go through AI by 2027 Comarketing-News, 12 October 2026
- Vodafone rolls out AI agents to more than 100,000 SMEs Economía Digital, 5 October 2026
- Joe AI raises €2 million to automate customer relations in real estate Maddyness, 1 October 2026
- Agentic AI: French companies held back by their CRM Relation Client Mag, 30 September 2026
- How Zeffy, the North American “cousin” of HelloAsso, is doubling in size every year thanks to AI Maddyness, 28 September 2026
- The AI Act and the Digital Omnibus: a postponement that exempts no one Journal du Net, 17 September 2026
- Katie Stein, CEO of ASAPP – Interview Series Unite.AI, 15 September 2026
- User trust is a barrier to integrating AI into the customer experience HayCanal, 14 September 2026
- AXA rolls out its AI agents: why insurance is changing scale Brief IA, 11 September 2026
- The automation ceiling and its negative impact on user experience iProfesional, 3 August 2026
Another question?
Claude searches the articles in the watch and answers you, citing its sources.
Answer written by an AI model (Claude, by Anthropic) from the articles in the watch only: check the figures in the sources. Questions are recorded;learn more.
Method
Text written on 20 October 2026 by an AI model (Claude Sonnet 5, by Anthropic) from 67 AI watch articles on this theme, published between 29 August 2025 and 19 October 2026. Each paragraph links to the entries it draws on, and every number in the text was found in one of these articles before publication. The text is rewritten every week.
Other industries
- AI in manufacturing
- AI at family offices and private wealth advisory firms
- AI in retail and distribution
- AI in transport and logistics
- AI in healthcare
- AI in consulting, law and accounting
- AI in software, IT services and telecoms
- AI in energy and utilities
- AI in construction and real estate
- AI in tourism and hospitality
- AI in media, communication and marketing
- AI in the public sector