Customer support specialists
For people who resolve customer questions and maintain service quality.
What is changing
The routine layer of the job is moving to machines first. Salesforce's State of Service: AI Agents Edition (a survey of 3,075 customer service professionals, reported 2026) found use of AI agents in service organizations rose from 39% in 2025 to 66% in 2026, and that 85% of service organizations now use at least one form of AI. The kinds of contacts being automated first are the repetitive ones — order status, password resets, easy returns (Klaviyo, June 29, 2026). Forrester research recapped by Destination CRM estimates AI can autonomously resolve 60–80% of routine customer inquiries, and cites Bank of America's Erica handling about 2 million transactions a day, offloading work of roughly 11,000 employees. Against that, Forrester's own blog (May 20, 2026) predicts AI will cause 49% of current customer service jobs to disappear by 2030.
But the aggregate data on actual headcount cuts is far more cautious. Gartner's October 2025 survey of 321 service leaders (recapped by Klaviyo) found 91% feel pressure to implement AI in 2026, yet only 20% had actually reduced agent headcount — 'the majority report that headcount remains steady, even as they support more customers.' The same survey found nearly 80% of organizations plan to move agents into new positions and 84% are adding new skills to frontline roles. Gartner has also warned of a rehiring cycle: across industries, 30% of employees laid off in AI-driven cuts will need to be rehired by 2029, and half of companies that reduced customer service headcount because of AI will rehire staff by 2027 to perform similar functions (both reported by CX Dive, 2026).
The content of remaining roles shifts upward. Forrester's Leggett and Ramos describe the core mandate moving 'from reactively interacting with customers directly to managing the AI that interacts with them': lower-tier reps manage teams of AI agents, unblock them on judgment calls, and feed the AI feedback; higher-tier reps specialize into technical subject-matter, policy-expert, and relationship roles; and new 'light technology' jobs appear for configuring agents and managing the quality of AI outputs (Forrester, May 20, 2026; same framing appears from Forrester's Max Ball in CX Dive — 'bot unblockers,' 'judges,' and 'experts'). The economics are still unsettled: Gartner's Emily Potosky told CX Dive that service and support spent a median of almost $6 million on AI in 2025, and of the use cases Gartner has evaluated, about a quarter delivered negative returns, about a quarter positive, and 42% yielded unclear value.
But the aggregate data on actual headcount cuts is far more cautious. Gartner's October 2025 survey of 321 service leaders (recapped by Klaviyo) found 91% feel pressure to implement AI in 2026, yet only 20% had actually reduced agent headcount — 'the majority report that headcount remains steady, even as they support more customers.' The same survey found nearly 80% of organizations plan to move agents into new positions and 84% are adding new skills to frontline roles. Gartner has also warned of a rehiring cycle: across industries, 30% of employees laid off in AI-driven cuts will need to be rehired by 2029, and half of companies that reduced customer service headcount because of AI will rehire staff by 2027 to perform similar functions (both reported by CX Dive, 2026).
The content of remaining roles shifts upward. Forrester's Leggett and Ramos describe the core mandate moving 'from reactively interacting with customers directly to managing the AI that interacts with them': lower-tier reps manage teams of AI agents, unblock them on judgment calls, and feed the AI feedback; higher-tier reps specialize into technical subject-matter, policy-expert, and relationship roles; and new 'light technology' jobs appear for configuring agents and managing the quality of AI outputs (Forrester, May 20, 2026; same framing appears from Forrester's Max Ball in CX Dive — 'bot unblockers,' 'judges,' and 'experts'). The economics are still unsettled: Gartner's Emily Potosky told CX Dive that service and support spent a median of almost $6 million on AI in 2025, and of the use cases Gartner has evaluated, about a quarter delivered negative returns, about a quarter positive, and 42% yielded unclear value.
Where people matter
Every source that describes what AI is good at also describes what it is not. Klaviyo's 2026 guidance puts it bluntly: 'AI breaks the moment a situation needs judgment, empathy, or cross-functional thinking' — AI can detect negative sentiment but not de-escalate an angry customer, handle scripted paths but break on edge cases, and generate templated apologies but not repair a brand relationship. Forrester's tier model reserves 'issues that require a human judgment call' for humans, either as bot unblockers at the frontline or as higher-tier 'judges' and 'experts' for problems too nuanced for AI (Forrester blog, May 20, 2026; Max Ball via CX Dive). The outcome data points the same way: Salesforce reports that customer satisfaction is the #1 KPI that improves after AI agents are deployed — ahead of agent productivity and handle time — which its analysts read as evidence that AI frees people for the work customers actually notice. Consumer sentiment is less enthusiastic than the technology: Klaviyo's 2026 consumer research found 41% of people say service chats that 'don't feel human' are the top too-automated experience, and Gartner's Potosky told CX Dive that 'people will never fully be replaced.'
Ways to adapt
Learn the tools while they are still scarce. Zendesk's CX Trends data shows only about one-fifth of agents currently have generative AI tools at their disposal — but 80% of employees who use AI say it has already improved the quality of their work. Agents who volunteer to pilot agent-assist and copilot features are building the exact skill Forrester says the job is becoming: managing and correcting AI output.
Close the training gap yourself rather than waiting for it. Zendesk found 72% of CX leaders believe they have provided adequate AI training, but 55% of agents say they have received none; only 45% of agents report any AI training and just 21% are satisfied with it. Only 34% of agents say they understand their department's AI strategy. Asking pointed questions about the roadmap — which tickets are being automated, what the escalation rules are — turns a knowledge gap into a visible asset.
Move toward the roles being created. Gartner's October 2025 survey found 58% of service leaders plan to upskill agents into knowledge management roles (Klaviyo), and Salesforce reports 72% of service operations professionals call data readiness a major blocker to AI — meaning people who can build and maintain the knowledge bases and clean data that AI depends on are in demand. Forrester adds that AI agent configuration will be done by 'citizen developers' with low-code tools, and that quality-management roles for AI outputs are appearing. In parallel, deepen the human work AI avoids: de-escalation, cross-system troubleshooting, and judgement on ambiguous cases (Klaviyo). With scarce training budgets, the practical route is a combination of self-directed practice with the company's AI tools, knowledge of the escalation design, and specialization in a domain the team's AI handles poorly.
Close the training gap yourself rather than waiting for it. Zendesk found 72% of CX leaders believe they have provided adequate AI training, but 55% of agents say they have received none; only 45% of agents report any AI training and just 21% are satisfied with it. Only 34% of agents say they understand their department's AI strategy. Asking pointed questions about the roadmap — which tickets are being automated, what the escalation rules are — turns a knowledge gap into a visible asset.
Move toward the roles being created. Gartner's October 2025 survey found 58% of service leaders plan to upskill agents into knowledge management roles (Klaviyo), and Salesforce reports 72% of service operations professionals call data readiness a major blocker to AI — meaning people who can build and maintain the knowledge bases and clean data that AI depends on are in demand. Forrester adds that AI agent configuration will be done by 'citizen developers' with low-code tools, and that quality-management roles for AI outputs are appearing. In parallel, deepen the human work AI avoids: de-escalation, cross-system troubleshooting, and judgement on ambiguous cases (Klaviyo). With scarce training budgets, the practical route is a combination of self-directed practice with the company's AI tools, knowledge of the escalation design, and specialization in a domain the team's AI handles poorly.
Adjacent paths
Knowledge management — building, maintaining, and curating the knowledge base that both human agents and AI agents draw on. Gartner's surveys show this is where many service leaders plan to move agents (58% per the October 2025 survey; Klaviyo). AI agent operations and configuration — setting up and tuning AI agents with low-code tools, which Forrester expects to become 'citizen developer' work (Forrester blog, May 20, 2026). Quality assurance for AI-powered service — reviewing AI conversations, scoring automated resolutions, and feeding corrections back; Forrester describes new jobs that 'manage the quality of AI outputs and processes.' Conversation and insights analysis — using conversation intelligence to surface product gaps, sentiment trends, and cost-per-intent data, which Forrester's insight teams already do. Escalation and specialist support — technical subject-matter expert, policy expert, and relationship-management roles that Forrester expects higher-tier reps to grow into. Business solutions architecture and related integration roles — which Gartner's Potosky told CX Dive companies will need to hire 'to keep AI solutions running and successful.' These are adjacent moves, not jump shifts: each builds directly on what a support specialist already does every day.
Try a small experiment
Write several fictional customer requests. Prepare responses with a tool, then check each against a small policy document you wrote. Include an ambiguous case that should be escalated to a person. Never use real customer data in this experiment.
Sources
- AI Will Reshape Customer Service Jobs In Dramatic Ways (Forrester Research (Kate Leggett and Laura Ramos), 2026-05-20)
https://www.forrester.com/blogs/ai-will-reshape-customer-service-jobs-in-dramatic-ways/
Supports: The shift of frontline reps from directly serving customers to managing AI agents, tiered specialization into technical/policy experts, new 'light technology' roles, and Forrester's forecast that AI could eliminate 49% of current customer service jobs by 2030.
- Are AI-driven customer service cuts here to stay? Gartner predicts rehiring (CX Dive (Informa TechTarget), )
https://www.customerexperiencedive.com/news/ai-driven-customer-service-cuts-gartner-predicts-rehiring/830114/
Supports: Gartner's predictions that 30% of workers laid off in AI-driven cuts will be rehired by 2029 and that half of companies cutting service headcount due to AI will rehire by 2027; Gartner ROI findings on service AI use cases; the Klarna rehiring case; Forrester analyst Max Ball's 'bot unblocker' and 'judges and experts' framing.
- New Research: AI Service Agents Improve Customer Satisfaction (Salesforce (State of Service: AI Agents Edition, n=3,075 service professionals), )
https://www.salesforce.com/news/stories/ai-service-agents-improve-customer-satisfaction/
Supports: AI agent adoption rising from 39% (2025) to 66% (2026), 85% of service organizations using some form of AI, customer satisfaction as the top improved KPI, 70% seeing value within 60 days, and data readiness as the most-cited blocker by service operations staff.
- Will AI replace customer service jobs? What the 2026 data says (Klaviyo, 2026-06-29)
https://www.klaviyo.com/blog/ai-customer-service-roles
Supports: Recap of Gartner's October 2025 survey (91% of leaders feel pressure to deploy AI, but only 20% actually cut agent headcount; ~80% plan to move agents into new roles; 84% adding new skills; 58% upskilling agents into knowledge management), plus Gartner's February/March 2026 rehiring and tech-spend predictions and Klaviyo consumer research on 'too automated' experiences.
- 59 AI customer service statistics for 2026 (Zendesk CX Trends data) (Zendesk, )
https://www.zendesk.com/blog/ai/productivity/ai-customer-service-statistics/
Supports: Agent-side gaps in AI adoption and training: only about one-fifth of agents have generative AI tools, 45% received AI training, 21% are satisfied with it, 55% of agents say they got no training while 72% of leaders believe they provided it, and only 34% of agents understand their department's AI strategy.
- AI Is Causing a Major Shift in Customer Service (Destination CRM (recap of Forrester labor-market research, by Phillip Britt), )
https://www.destinationcrm.com/Articles/CRM-Insights/Insight/AI-Is-Causing-a-Major-Shift-in-Customer-Service-176520.aspx
Supports: Customer service job postings already 10% below pre-pandemic levels, stagnating CSR pay since mid-2025, Forrester's estimate that AI can autonomously resolve 60–80% of routine inquiries, and the Bank of America Erica example (about 2 million transactions a day, offloading work of roughly 11,000 employees).
https://www.forrester.com/blogs/ai-will-reshape-customer-service-jobs-in-dramatic-ways/
Supports: The shift of frontline reps from directly serving customers to managing AI agents, tiered specialization into technical/policy experts, new 'light technology' roles, and Forrester's forecast that AI could eliminate 49% of current customer service jobs by 2030.
- Are AI-driven customer service cuts here to stay? Gartner predicts rehiring (CX Dive (Informa TechTarget), )
https://www.customerexperiencedive.com/news/ai-driven-customer-service-cuts-gartner-predicts-rehiring/830114/
Supports: Gartner's predictions that 30% of workers laid off in AI-driven cuts will be rehired by 2029 and that half of companies cutting service headcount due to AI will rehire by 2027; Gartner ROI findings on service AI use cases; the Klarna rehiring case; Forrester analyst Max Ball's 'bot unblocker' and 'judges and experts' framing.
- New Research: AI Service Agents Improve Customer Satisfaction (Salesforce (State of Service: AI Agents Edition, n=3,075 service professionals), )
https://www.salesforce.com/news/stories/ai-service-agents-improve-customer-satisfaction/
Supports: AI agent adoption rising from 39% (2025) to 66% (2026), 85% of service organizations using some form of AI, customer satisfaction as the top improved KPI, 70% seeing value within 60 days, and data readiness as the most-cited blocker by service operations staff.
- Will AI replace customer service jobs? What the 2026 data says (Klaviyo, 2026-06-29)
https://www.klaviyo.com/blog/ai-customer-service-roles
Supports: Recap of Gartner's October 2025 survey (91% of leaders feel pressure to deploy AI, but only 20% actually cut agent headcount; ~80% plan to move agents into new roles; 84% adding new skills; 58% upskilling agents into knowledge management), plus Gartner's February/March 2026 rehiring and tech-spend predictions and Klaviyo consumer research on 'too automated' experiences.
- 59 AI customer service statistics for 2026 (Zendesk CX Trends data) (Zendesk, )
https://www.zendesk.com/blog/ai/productivity/ai-customer-service-statistics/
Supports: Agent-side gaps in AI adoption and training: only about one-fifth of agents have generative AI tools, 45% received AI training, 21% are satisfied with it, 55% of agents say they got no training while 72% of leaders believe they provided it, and only 34% of agents understand their department's AI strategy.
- AI Is Causing a Major Shift in Customer Service (Destination CRM (recap of Forrester labor-market research, by Phillip Britt), )
https://www.destinationcrm.com/Articles/CRM-Insights/Insight/AI-Is-Causing-a-Major-Shift-in-Customer-Service-176520.aspx
Supports: Customer service job postings already 10% below pre-pandemic levels, stagnating CSR pay since mid-2025, Forrester's estimate that AI can autonomously resolve 60–80% of routine inquiries, and the Bank of America Erica example (about 2 million transactions a day, offloading work of roughly 11,000 employees).