Will AI replace call center jobs? What is changing, and what to learn next
Some of it, yes. Chatbots and AI agent-assist tools already handle password resets, order tracking and other simple contacts, and US government projections show customer service rep jobs shrinking over the next ten years. That is worth taking seriously.
The work that stays human is the harder work: escalations, complaints, retention and sales, and the people who check, train and supervise the AI itself. Below are the skills that move you toward that work, with a first step for each.
Talk to Carmen
Your AI career coach
Say it the way you’d say it to a friend. Carmen will ask what she needs to know.
5% decline
Projected change in US customer service rep jobs, 2025 to 2035
BLS Occupational Outlook HandbookAgentic AI resolves 80% of common customer service issues without a human
Gartner prediction (March 2025) for 2029
Gartner press release, March 5, 202550% of organizations that expected to significantly reduce their customer service workforce abandon those plans
Gartner prediction (June 2025) for 2027
Gartner press release, June 10, 2025Skills to learn next
Handling escalations and complaints
When a bot cannot solve something, the customer arrives angrier and the problem is harder. Agents who can calm people down and fix complicated cases become more valuable as simple contacts disappear.
- What it takes
- Free to practice. Ask to take escalated calls or shadow your escalations desk.
- First step
- Ask your TL if you can shadow the escalations or supervisor line for one shift.
Retention and sales
Saving a customer who wants to cancel, or upselling at the right moment, depends on reading the person. Companies protect these roles because they bring in revenue.
- What it takes
- Often an internal move to a retention or sales queue.
- First step
- Listen to three of the best retention or sales calls on your account and write down the exact phrases that worked.
Working well with AI agent-assist tools
Many accounts now use tools that suggest answers, summarize calls or write notes. Agents who check the suggestions, catch the wrong ones, and give feedback become the go-to people when new tools roll out.
- What it takes
- Free. You learn it on the job.
- First step
- Keep a list this week of every time the AI suggestion on your account was wrong, and share it with your TL.
QA and call analytics
Someone has to review what both humans and bots say to customers, spot patterns, and report them. QA roles are moving toward reviewing AI conversations as well as agent calls.
- What it takes
- Internal move in most BPOs. Strong QA scores of your own help.
- First step
- Ask your QA team how they review bot or chatbot conversations, and whether they need help.
CX tool administration (Zendesk, Freshdesk, Salesforce Service Cloud)
Setting up the ticketing system itself: macros, routing rules, help center articles, and the bot's answers. Suits agents who already know the tool better than the people who configured it.
- What it takes
- Zendesk and Salesforce both offer free learning material. Formal certifications cost money, so check job postings first.
- First step
- Find which ticketing tool your account uses and complete its free beginner admin course.
Excel and basic SQL for WFM and reporting
Forecasting, scheduling and reporting all run on data. Strong Excel opens the door to WFM and reporting analyst roles, and basic SQL takes you further.
- What it takes
- Free online courses cover the basics. Plan for a few weeks of regular practice.
- First step
- Download last week's metrics for your team (if allowed) and build one pivot table showing AHT by agent.
Knowledge base and bot content writing
Chatbots are only as good as the answers they are given. People who write clear help articles and bot replies, and fix them when customers get stuck, are needed on every account that uses AI.
- What it takes
- Free to start. Good written English and knowing the product are the main requirements.
- First step
- Pick one question customers ask you every day and rewrite the help article for it so a customer could solve it alone.
Sample conversation
How a conversation with Carmen starts
Carmen asks before she suggests. An illustrative example, not a real person.
Our account just added a chatbot and half my easy tickets are gone. Am I next?
That is a fair worry. The tickets left for you now are the harder ones, which is where people are still needed. What do you usually handle now that the bot cannot?
Billing disputes, and people who already tried the bot and are angry.
Those are escalation skills, and they grow in value. Want to look at a path toward QA or an escalations desk on your account?
Work it out with Carmen, free
30 free minutes with Carmen and the team. No card needed.
Questions & answers
- Will AI replace call center jobs?
- It is replacing some of the work, mostly simple, repetitive contacts. US projections show fewer customer service rep jobs over the next decade, but still many openings each year. Complex, emotional and revenue-related work, and the jobs that supervise AI, are more likely to stay.
- Is the BPO industry in the Philippines shrinking because of AI?
- According to IBPAP figures reported in January 2026, the Philippine IT-BPM industry grew in 2025, to about 1.9 million workers. The mix of work is changing, so building skills beyond simple voice and chat support is a sensible move.
- Which call center jobs are safest from AI?
- Escalations, retention, sales, QA, training, workforce management, and roles that set up and supervise AI tools. These rely on judgment, trust and people skills.
- Should I learn AI tools as a call center agent?
- Yes, starting with the ones your account already uses. Knowing where they go wrong makes you useful when your company rolls out the next one.
More for call center and bpo agents
Learn a new skill
Sources
- BLS Occupational Outlook Handbook: Customer Service Representatives
- Gartner press release, March 5, 2025: Agentic AI will autonomously resolve 80% of common customer service issues by 2029
- Gartner press release, June 10, 2025: 50% of organizations will abandon plans to reduce customer service workforce due to AI
- The Philippine Star: IT-BPM revenues breach $40 billion mark (IBPAP figures, January 2026)
Last checked September 29, 2026.