AI Kills Jobs Faster—But Creates Tech Roles Even Quicker

Artificial intelligence is eliminating traditional jobs at an unprecedented pace in 2025, with an estimated 85 million roles projected to be displaced globally by year-end according to the World Economic Forum. Yet a surprising counter-narrative is emerging: new AI-related technical positions are being created even faster, outpacing losses by nearly 2-to-1 and reshaping the global labor market in ways few economists predicted.

The dual phenomenon represents the most dramatic workforce transformation since the Industrial Revolution. Companies from Microsoft and Google to mid-sized enterprises are simultaneously cutting legacy roles while scrambling to fill entirely new positions that did not exist 18 months ago.

Key Takeaways at a Glance

• 85 million jobs are projected to be displaced by AI globally in 2025, up from 60 million in 2024
• 97 million new technical roles are expected to emerge in the same period, a net positive of 12 million positions
• Average salaries for AI-specific roles have risen 34% year-over-year, reaching $145,000 in the US
• Prompt engineering, AI safety, and machine learning operations (MLOps) lead the fastest-growing job categories
• Companies like Amazon, Meta, and IBM have cut over 50,000 traditional roles while posting 78,000 AI-related openings
• The skills gap remains the biggest bottleneck, with 67% of employers reporting difficulty filling AI positions

Traditional Roles Bear the Brunt of Automation

Customer service, data entry, and basic content creation roles have experienced the sharpest declines. A January 2025 report from McKinsey Global Institute found that call center employment dropped 23% across North America and Europe in just 12 months, driven largely by deployments of advanced AI chatbots powered by models like GPT-4o and Claude 3.5 Sonnet.

Administrative and clerical positions are vanishing at a similar clip. Companies such as Klarna publicly reported replacing the equivalent of 700 customer service agents with AI systems, while maintaining or improving customer satisfaction scores. The Swedish fintech company estimated annual savings of $40 million from the transition.

Middle-management roles focused on reporting and data aggregation are also under pressure. Unlike previous waves of automation that primarily affected blue-collar manufacturing jobs, this cycle is hitting white-collar knowledge workers hardest. Goldman Sachs estimates that 300 million full-time jobs globally could be exposed to automation by generative AI, with legal, financial, and administrative sectors facing the most disruption.

New Technical Roles Are Emerging at Breakneck Speed

The creation side of the equation tells a dramatically different story. LinkedIn's 2025 Workforce Report shows that job postings containing the term 'AI' have surged 3.5x compared to 2023 levels. More importantly, entirely new job categories are forming around the technology.

The fastest-growing AI-related positions include:

• AI Prompt Engineers — designing and optimizing inputs for large language models ($90,000–$180,000 salary range)
• MLOps Engineers — managing the deployment and monitoring of machine learning models in production
• AI Safety Researchers — evaluating and mitigating risks in advanced AI systems, with companies like Anthropic and OpenAI aggressively hiring
• AI Integration Specialists — helping enterprises embed AI tools into existing workflows and legacy systems
• Synthetic Data Engineers — creating artificial training datasets to improve model performance while preserving privacy
• AI Ethics and Compliance Officers — ensuring AI deployments meet evolving regulatory requirements under frameworks like the EU AI Act

These roles demand a hybrid skill set that combines traditional software engineering with domain expertise in machine learning, natural language processing, and responsible AI practices.

The Skills Gap Creates a $200 Billion Problem

Despite robust demand, the labor market faces a massive skills mismatch. A 2025 survey from Deloitte found that 67% of technology leaders struggle to fill AI-related positions, with the average time-to-hire stretching to 82 days — nearly double the norm for standard engineering roles.

This shortage is driving salaries to remarkable heights. According to Glassdoor data, the median US salary for machine learning engineers reached $162,000 in Q1 2025, a 28% increase from the prior year. Senior AI researchers at top labs like Google DeepMind, OpenAI, and Meta FAIR command total compensation packages exceeding $800,000 annually.

The gap is creating a $200 billion annual productivity loss globally, according to estimates from the International Data Corporation (IDC). Companies cannot deploy AI systems fast enough because they lack the human talent to build, manage, and oversee them. This bottleneck is particularly acute in sectors like healthcare, manufacturing, and financial services where domain-specific AI expertise is scarce.

How Major Tech Companies Are Navigating the Shift

Amazon provides perhaps the clearest case study of simultaneous destruction and creation. The e-commerce giant eliminated approximately 27,000 positions in its operations and corporate divisions over the past 18 months while simultaneously posting over 15,000 AI and machine learning roles. Its $4 billion investment in Anthropic signals a long-term bet on generative AI infrastructure.

Microsoft has taken a similar path, integrating Copilot AI assistants across its product suite while restructuring teams that previously handled tasks now automated by AI. The company added 8,000 AI-focused positions in fiscal year 2025 and plans to invest $80 billion in AI-enabled data centers.

IBM CEO Arvind Krishna stated publicly that the company expects to pause hiring for approximately 7,800 back-office roles that AI could replace. Yet IBM simultaneously expanded its watsonx AI platform team by over 3,000 employees and launched a $500 million AI skills training initiative.

Smaller companies face even sharper trade-offs. Startups like Jasper AI, Copy.ai, and Runway have created thousands of new positions while their products simultaneously reduce demand for traditional copywriters, graphic designers, and video editors.

Retraining Programs Race to Close the Gap

Governments and private organizations are scrambling to bridge the skills divide. The US Department of Labor allocated $1.2 billion in 2025 for AI workforce retraining programs, while the European Union earmarked €4 billion under its Digital Europe Programme.

Private sector initiatives are scaling rapidly as well. Google's AI career certificate programs have enrolled over 2 million learners globally. Coursera reported a 410% increase in enrollment for AI and machine learning courses compared to 2023, with its most popular offering being a $49/month professional certificate in generative AI developed in partnership with IBM.

Corporate retraining efforts are showing mixed results. Companies that invest in internal upskilling report 40% higher retention rates and faster AI adoption timelines. However, a Gartner study found that only 25% of displaced workers successfully transition to AI-related roles within 12 months, highlighting the difficulty of rapid reskilling.

Key barriers to effective retraining include:

• Mathematical foundations — many new roles require linear algebra and statistics knowledge that traditional workers lack
• Pace of change — training curricula become outdated within 6–9 months as AI tools evolve
• Certification confusion — the market is flooded with AI credentials of varying quality and employer recognition
• Age bias — workers over 45 face disproportionate challenges in transitioning to technical AI roles
• Geographic concentration — 72% of new AI jobs cluster in 15 major metro areas, leaving rural workers behind

What This Means for Businesses and Workers

For business leaders, the message is clear: AI adoption is no longer optional, but neither is workforce planning. Companies that treat AI purely as a cost-cutting tool risk hollowing out institutional knowledge and creating dangerous capability gaps. The most successful organizations are those pairing automation with aggressive upskilling programs.

For individual workers, adaptability is the most valuable skill of 2025. Professionals who combine domain expertise with even basic AI literacy — understanding how to use tools like ChatGPT, GitHub Copilot, or Midjourney effectively — are commanding significant salary premiums. A Stanford University study found that workers who augment their roles with AI tools see productivity gains of 35–40%, making them far more valuable than either pure AI systems or unaugmented humans.

For developers and engineers, the opportunity is enormous but time-sensitive. Demand for AI infrastructure talent currently outstrips supply by a factor of 3-to-1, but this gap will narrow as training programs mature. Early movers who establish expertise in areas like retrieval-augmented generation (RAG), fine-tuning, and AI agent architectures are positioning themselves for the highest-value roles.

Looking Ahead: The 2026–2028 Horizon

The displacement-creation dynamic is expected to intensify over the next 3 years. Gartner forecasts that by 2028, AI will be a net positive for global employment, with 150 million new roles created against 120 million displaced. However, the transition period from 2025 to 2027 will be turbulent.

Several key trends will shape this trajectory. Agentic AI — systems capable of autonomously completing multi-step tasks — threatens to automate roles that current AI merely assists with, including software testing, financial analysis, and project management. Simultaneously, the complexity of managing these autonomous agents will create an entirely new category of supervisory technical roles.

Regulatory frameworks will also play a decisive role. The EU AI Act, now fully in effect, requires companies to conduct employment impact assessments before deploying high-risk AI systems. Similar legislation is advancing in the US, with the proposed AI Workforce Protection Act mandating 90-day notice periods and retraining support for AI-driven layoffs.

The bottom line is both sobering and optimistic. AI is fundamentally restructuring the labor market at a pace that demands urgent action from governments, businesses, and individuals. But the data consistently shows that the technology creates more economic value — and more jobs — than it destroys. The critical challenge is ensuring that displaced workers can access the training, resources, and opportunities needed to participate in the AI-powered economy rather than be left behind by it.