Job Displacement | By 2028
Customer Service Automation by 2028
Weighted average across 6 sources. Observed so far: ~42% (4 measurements from Yale Budget Lab, Brookings, Dallas Fed, BLS). Projections range 25–55% (median ~40%).
41.2% of all customer service interactions (phone calls, chat messages, emails) will likely be handled by AI without a human agent ever getting involved. This doesn't mean 41.2% of CS jobs disappear, since the remaining interactions may need more skilled human agents, but headcount reductions are widely expected. Unlike software or creative work, customer service demand is relatively inelastic: consumers don't want more support interactions at lower prices. This makes displacement more likely than demand expansion in this sector.
Blended estimate across 6 sources ranging 12–66%. Higher-tier evidence and more recent data are weighted more heavily. See the full methodology for details on weighting, source validity, and recency bias.
Observed Data & Projections
This prediction has two fundamentally different types of evidence: observed employment data (what has actually happened) and forward-looking projections (what researchers estimate will happen). They are shown separately below because they answer different questions.
Filter by evidence tiers
Note: Measures interaction automation, not job elimination. Actual headcount effects depend on demand response and service quality. A high automation rate may reduce headcount but does not imply equivalent job loss.
What has happened
Measured employment data from government statistics, large-scale surveys, and administrative records. This is ground truth: what has actually occurred in the labor market.
Each dot is a different measurement source. Click any dot to jump to its source below.
What researchers project
Forward-looking estimates from structural models, institutional surveys, and expert forecasts. All projections target by 2028, shown by the reference line. The wide range (25–55%) reflects different model assumptions about reinstatement effects, demand elasticity, and adoption speed, not just parameter uncertainty.
Each dot is a different projection source. The x-axis shows when the projection was published. Click any dot to jump to its source. Overlay bars show directional signals from related studies.
Task Visualizer
What parts of your job will be cheaper to do with AI?
See which of your tasks face cost pressure from AI first.
Full Economy Picture
AI and the US Economy
Automation impact by occupation and income tier.
Sources (34)
Brookings: CS rep a key Gateway role; AI disruption cascades
if AI were to significantly reorient or automate a customer service role, the opportunity for economic mobility for workers in Origin roles such as receptionists and clerks would likely also be impaired
Stanford DEL: 82% ticket deflection at tech co via AI-first customer support
90 or 95% are now fully automated by an agent. If someone says their food didn't arrive or something went wrong with their order, 90 to 95% of those are completely automated. In customer support, a technology company achieved 82% ticket deflection by redesigning workflows around AI first resolution.
FRI: Rapid scenario → AI surpasses all human customer service agents by 2030
In the rapid AI scenario (14% probability), AI systems surpass all freelance customer service agents, handling inquiries as inexpensively and reliably as humans.
Anthropic: CS rep tasks prevalent in automated API workflows
customer service tasks, including, for example, automated support for payment and billing issues, are prevalent in the API data. These contributed to a higher observed exposure for Customer Service Representatives—Claude was recorded doing a high share of their tasks in automated workflows, so these jobs may be more likely to change as AI diffuses.
Lodefalk et al.: Sweden customer service shows same age-gradient AI displacement pattern
Customer service agents (SSYK 4222) display the same age gradient as the economy-wide canaries effect, lying between the augmentation and automation poles. All four spotlight occupations show young workers declining, older workers stable or rising.
Fed/Duke: Office/admin support highest negative AI exposure index (NEI 2.0)
Office and administrative support roles exhibit the most negative AI exposure (NEI 2.025), consistent with automation of routine clerical activities such as data entry. Routine clerical employment share expected to decline by 0.76pp in 2026 and 2.19pp by 2028.
Anthropic: Customer service reps 70.1% observed AI task exposure, rising in API traffic
Customer service representatives at 70.1% observed exposure, whose main tasks we increasingly see in first-party API traffic.
KPMG: 73% automating workflows spanning functions
73% Automating workflows that span multiple functions
Yale Budget Lab: Business support services score 8.55/10 exposure (highest industry); CS reps score 9/10
Business support services (NAICS 561400) has the highest weighted industry exposure at 8.55/10, with customer service representatives at 9/10 comprising 53.8% of sector employment. Office/admin SOC group: mean 7.70/10 (range 4-9).
Gartner: ~80% plan to transition CS agents to new roles, 84% expanding skills
Only 20% of customer service leaders have actually reduced agent staffing due to AI, yet 80%+ expect to reduce headcount over the next 18 months. Nearly 80% plan to transition agents into new roles. Survey of 321 CS leaders (Sep-Oct 2025).
Deutsche Bank: Up to 75% of CS interactions could be automated by 2026
Customer service could see as much as 75% of interactions automated by 2026, reserving human agents for complex, nuanced cases.
Brookings: CS reps face high AI exposure + low adaptive capacity
Customer service representatives get the short end of the stick in both categories [high AI exposure and low adaptive capacity].
95% of customer interactions projected to be AI-powered by 2028. Current AI resolution rates at leading companies exceed 55% without human escalation.
MGI: Advanced AI agents could eventually handle 80-90% of customer inquiries
Together, these now handle roughly 40 percent of all calls, resolving more than 80 percent without human involvement. The new process has cut the average cost per call by about 50 percent. Advanced AI agents could eventually handle 80 to 90 percent of customer inquiries.
Gartner: 20-30% of CS agents replaced by 2026
Gartner projects 20-30% of customer service agents will be replaced by generative AI by end of 2026. AI adoption in CS jumped from 46% to 61% in one year.
BofA Erica: 2B interactions, 98% resolution rate
BofA's Erica AI assistant surpassed 2 billion client interactions with a 98% resolution rate. Now handles 56% of all customer inquiries without human escalation.
E-commerce field experiment: AI chatbots increased sales 16%
AI chatbots for customer service increased sales by 16%; AI-generated product descriptions increased sales by 2.05% in field experiments.
Salesforce: Cut CS workforce 44% (9K to 5K); AI agents handle 50% of conversations
50% are with agents, 50% are with humans. From 9,000 heads to about 5,000 because I need less heads. Support costs reduced by 17%.
Stanford: 11% decline for early-career CS workers
~11% decline for early-career customer service workers; declines concentrated in roles where AI automates rather than augments tasks.
Klarna reversed course, rehiring humans after 22% CSAT drop
Klarna reversed its AI-only CS strategy after customer satisfaction dropped 22%. CEO Siemiatkowski: 'we went too far.' Now operates hybrid model with AI handling ~66% but humans for complex cases.
Cisco: 68% of CS interactions by agentic AI by 2028
Survey of 7,950 decision-makers: 68% of all CS interactions expected to be handled by agentic AI by 2028. 56% expected within 12 months.
ILO-NASK: Clerical support most exposed; data entry/payroll highest automation potential
Clerical support roles remain the most exposed. Data entry clerks, payroll clerks, and typists continue to show the highest automation potential.
QJE: 15% productivity gain, novices benefit most
AI assistance increases customer service worker productivity by 15% on average. Less experienced workers improve most; AI disseminates best practices from top performers.
Gartner: agentic AI to resolve 80% of common CS issues by 2029
Agentic AI will autonomously resolve 80% of common customer service issues by 2029. Currently only 1.6% of interactions are fully automated end-to-end.
Teleperformance: 200+ new AI projects launched; investing EUR 100M in AI partnerships
Teleperformance reported full-year 2024 revenue of EUR 10,280 million. Plans to invest up to EUR 100 million in AI-focused partnerships during 2025. Over 200 new AI projects launched in 2024.
Shopify reported its AI assistant Sidekick now resolves 40% of merchant support queries without human intervention, up from 12% one year prior.
BLS: CS representative employment projected to decline 5% through 2033
Customer service representatives, whose employment is projected to decline by 5.0 percent through 2033.
Gartner: By 2027, 25% of orgs will use chatbots as primary CS channel
Indeed: Customer service postings declined 18% YoY; now require AI tool familiarity
Customer service representative postings declined 18% year-over-year. Remaining postings increasingly require technical skills and AI tool familiarity.
NBER: AI boosts CS worker productivity 14%
AI assistance increased worker productivity by 14% on average, with greatest gains for novice and low-skilled workers.
HBR: GenAI will enhance, not erase, customer service jobs
Klarna's AI assistant handled two-thirds of customer service chats within one month of launch — 2.3 million conversations, doing the equivalent work of 700 full-time agents.
Gartner: Conversational AI will reduce contact center labor costs by $80B in 2026
Forrester: Report on the state of customer service automation
Know a study we’re missing? Suggest a source for this prediction.