— In New York and beyond, companies are increasingly using artificial intelligence to handle routine call-center work; while AI speeds up simple fixes and supplies agents with richer customer data, human representatives remain necessary for complex, sensitive or ambiguous problems, industry workers and executives say.
Key takeaways
- About 3 million Americans work in call-center roles; many more work globally answering billions of customer inquiries each year.
- AI tools now automate routine “break/fix” tasks and surface customer profiles, reducing time spent on menus and manual note-taking.
- Some firms cut large teams after adopting chatbots — Klarna replaced a 700-person department in 2023 — but later rehired staff when AI struggled with complex cases such as identity theft.
- Major providers such as Bank of America report broad chatbot use (Erica has been used billions of times since 2018) while vendors like Replicant and OpenAI push more capable conversational agents.
- Policy measures, like the bipartisan “Keep Call Centers in America Act,” aim to preserve human access to agents and incentivize domestic job retention.
Verified facts
Armen Kirakosian, 29, who works for TTEC and lives in Athens, Greece, recalls that his first call-center job required heavy menu navigation and manual note-taking. Today, Kirakosian says AI presents full customer profiles on screen and often predicts issues before the caller speaks, allowing him to devote more time to solving the problem.
Roughly 3 million Americans are employed in call-center positions, and many more worldwide handle inquiries spanning products and services, from electronics repairs to banking and retail orders. TTEC provides outsourced customer service across 22 countries, serving industries that rely on scalable contact-center capacity.
Industry studies and vendor forecasts differ widely on job impacts. Some analysts project modest single-digit percentage job losses, while other estimates foresee larger reductions over a decade. McKinsey has noted that attrition is high—about half of customer service agents leave within a year—partly because of stress and repetitive work.
Klarna replaced a 700-person customer-service team with AI systems in 2023 and later rehired a small number of agents when customer satisfaction fell on complex issues, including identity-theft cases. Bank of America says its chatbot Erica, launched in 2018, has been used roughly 3 billion times and now routes or escalates queries it cannot resolve.
Context & impact
Call centers have long used interactive voice response (IVR) menus to route calls, but customers often press “0” to reach a person when menus are frustrating or unclear. New AI-driven routing aims to interpret natural-language requests and send callers directly to the right resource without labyrinthine menus.
Vendors such as OpenAI are introducing products billed as conversational agents for booking and research tasks; firms like Replicant develop AI that mimics human conversational patterns. Businesses that succeed tend to combine automation for routine issues with well-trained human teams for complex, emotional or high-risk interactions.
For companies, AI can cut handling time and reduce costs on predictable requests, while human agents are increasingly focused on dispute resolution, fraud investigations, empathy-driven conversations and cases that require judgment across incomplete information.
Policymakers have taken notice: a bipartisan bill, the Keep Call Centers in America Act, would require clear ways for customers to reach a human and provide incentives for companies that maintain U.S.-based call-center jobs.
AI is enabling an “AI-first” model where bots manage many conversations and fewer, better-trained humans handle the hardest cases, accelerating a shift in workforce skills and pay structure.
Replicant / industry representative (paraphrased)
Unconfirmed
- Long-term net employment effects remain uncertain; estimates vary widely and depend on adoption pace, regulation and new service lines.
- Predictions that up to half of call-center jobs will disappear in a decade are projections, not settled outcomes.
Bottom line
AI is already changing how contact centers operate: routine tasks and routing are becoming automated, and agents have richer information at their fingertips. But evidence from major firms shows AI alone does not solve every problem; companies that blend automation with trained human staff tend to preserve customer satisfaction while reconfiguring job roles and training needs.