Is YOUR specific job at risk from AI?
Free 60-second check. No signup. Goldman Sachs + McKinsey + WEF 2026 data.
Check My Job Risk FreeThese Are Not Projections — These Already Happened
The AI job replacement conversation is dominated by predictions and statistics. This post is different: these are documented cases of specific jobs being replaced by AI in 2024–2025, with information about what happened to the workers involved.
Case 1: 700 Klarna Customer Service Agents
What happened: In February 2024, Klarna announced its AI assistant had performed the work of 700 customer service agents in its first month. The company reduced headcount from approximately 5,000 to 3,800 over 2022-2024.
The specific roles replaced: Customer service representatives handling standard account queries, dispute resolution for routine cases, payment plan inquiries, and order status questions.
What happened to the workers: Klarna's reduction was handled primarily through attrition — not filling positions as people left — rather than mass layoffs. Some workers transitioned to more complex customer escalation roles. The company stated it was investing in higher-skilled service roles while reducing headcount in standard service functions.
The lesson: The transition from 700 agent-equivalents happened without a single high-profile "firing" announcement. The positions simply did not exist when people moved on.
Case 2: BT Group — 55,000 Jobs Including 10,000 AI Replacements
What happened: BT Group (British Telecom) announced in May 2023 that it would reduce its workforce by 55,000 by the end of the decade, with AI replacing up to 10,000 of those positions.
The specific roles replaced: Network operations monitoring (AI can detect and respond to network faults faster), customer service (AI handling routine account management), back-office administration (AI processing billing, contracts, and compliance documentation).
What happened to the workers: BT offered retraining programs for affected employees. Unions negotiated extended notice periods and enhanced severance. Many workers moved into adjacent roles in network management and AI supervision. The timeline is gradual — the full 55,000 reduction targets 2030.
The lesson: Large employers are planning AI workforce reductions on 5-7 year timelines. Workers in affected roles have time to transition IF they start now.
Case 3: IBM — 7,800 Back-Office Roles
What happened: IBM CEO Arvind Krishna announced in May 2023 that IBM was pausing hiring in functions that AI could replace, targeting 7,800 jobs over five years. The company simultaneously invested in AI development and hired AI engineers.
The specific roles replaced: HR functions (benefits processing, employee query responses, documentation), finance operations (expense processing, accounts payable automation), and administrative support roles.
What happened to the workers: IBM offered career transition support, retraining programs, and in some cases reassignment to AI-adjacent roles. The company framed the shift as redeployment, though net headcount in affected functions is declining.
The lesson: Even a company deeply invested in AI and acknowledging the employment impact offered transition support — suggesting workers in affected roles at IBM had meaningful time to prepare.
Case 4: Duolingo Contractor Cuts
What happened: In January 2024, Duolingo laid off approximately 10% of its contractor workforce. CEO Luis von Ahn explicitly stated the reason: AI could now produce the content those contractors had been creating.
The specific roles replaced: Content creators producing exercises and lessons, translators doing quality checks on AI-generated translations, localization specialists, and curriculum designers for standard content types.
What happened to the workers: As independent contractors, they received less transition support than full employees. Many have reported difficulty finding equivalent work, as similar content roles at other companies are facing the same AI pressure simultaneously.
The lesson: Contractor and freelance roles face the fastest AI displacement with the least protection. Workers in gig content creation roles are among the most immediately exposed.
Case 5: Deutsche Bank and Financial Services Broadly
What happened: Deutsche Bank has reduced headcount by approximately 18,000 over 2019-2024. While restructuring has multiple causes, the bank has explicitly cited AI and automation as enabling significant operations reductions in back-office and middle-office functions.
The specific roles replaced: Trade settlement confirmation, loan documentation processing, compliance document review, basic risk reporting, and customer onboarding document handling.
What happened to the workers: Major European banks have offered extended transition periods, retraining through programs like Deutsche Bank's "db Professional Plus," and in some cases early retirement packages. Some workers transitioned to AI oversight and quality control roles.
Upload Resume - Get Free Risk Score
The Consistent Pattern Across All Cases
Every case study follows the same arc:
- Routine cognitive tasks in a defined domain are targeted first
- The transition happens through attrition, not mass firings
- The announcement (if any) comes well before actual displacement
- Workers with time to prepare achieve significantly better outcomes
- New AI-adjacent roles are created — but require different skills
What This Means for Your Planning
The most useful thing these case studies show is the timeline pattern: companies are working on 3-7 year automation roadmaps. The IBM announcement in 2023 targets 2028. BT's 2023 announcement targets 2030. Workers in those roles have years, not months.
But that window requires using the time. Workers at Klarna who moved into complex escalation roles before the AI cutover maintained employment. Workers in routine roles who waited for explicit signals found a shrunken job market.
Frequently Asked Questions
Are these AI job replacements typical or exceptions? They are representative of a broad pattern, not outliers. IBM, BT, Klarna, Deutsche Bank, and Duolingo are large enough that their cases are documented — but the same attrition-based AI replacement is happening at thousands of smaller companies without press coverage.
How much warning do workers typically get? Major companies typically announce automation plans 3-5 years in advance, giving workers meaningful transition time IF they act on the signals. Contractor and gig workers receive the least warning — often none.
My company has not announced any AI plans — does that mean I am safe? Not necessarily. Most companies do not make public announcements about automation roadmaps even when actively implementing them. The absence of an announcement is not evidence of safety — use the free calculator to assess your role's exposure independently.
How do I find out if my role is likely to be automated? Upload your resume at JobReplacementAI.com for a personalized assessment based on your actual tasks and skills. The calculator identifies which specific parts of your role match the automation patterns documented in cases like these — in 60 seconds, free.