Share and intensity of work current AI systems can materially affect.
Software Developers AI displacement risk
Code generation changes the junior task bundle, but architecture, debugging, security, product judgment, and system ownership keep the role augmentation-heavy.
Likely potential for exposed tasks to move to software after workflow integration.
Displacement risk is uneven inside the occupation. Entry-level work and boilerplate-heavy teams may change faster than ownership-heavy engineering roles.
Task profile
Where AI changes the work
Generate boilerplate
Exposure 83, automation 57%, augmentation 76%.
Write tests
Exposure 66, automation 34%, augmentation 71%.
Debug production issues
Exposure 38, automation 13%, augmentation 62%.
Design architecture
Exposure 22, automation 5%, augmentation 44%.
Transition pathways
Adjacent moves that preserve existing skills
role redesign
Low AI Platform Engineer
Training horizon: 4-9 months. Skill overlap 76. Wage preservation signal 112.
- Ship internal AI tooling
- Evaluate model failure modes
- Own secure deployment patterns
adjacent role
Low Security Automation Engineer
Training horizon: 6-12 months. Skill overlap 61. Wage preservation signal 108.
- Automate code scanning
- Review AI-generated patches
- Study threat modeling
Labor-market context
Median wage: $132,270. Employment context: High wage, high augmentation exposure. Typical education: Bachelor's degree common.
- Junior task bundle is exposed
- Demand favors product-aware builders
- Tool fluency matters
Sources
Evidence trail
McKinsey Global Institute / 2023 Generative AI and the Future of Work in America OECD / 2023 Employment Outlook 2023: Artificial Intelligence and the Labour Market O*NET / current public database Occupation, task, skill, and work-activity data US Bureau of Labor Statistics / current public release Occupational Employment and Wage Statistics