Will my job exist?

Industry forecast

Will AI replace tech & software jobs?

Every tech & software role we've scored, from most to least exposed to AI.

24 jobs · average risk 62%

  1. 01Manual Software TesterTest generation, execution and triage are among the fastest-automating tasks in software.87%
  2. 02IT Support TechnicianSelf-healing endpoints and AI helpdesks absorb most tier-1 tickets.81%
  3. 03Software Quality Assurance AnalystInterpreting QA as software quality assurance, routine testing is highly exposed to automation, with 2036 a speculative midpoint for majority displacement rather than a firm deadline.80%
  4. 04Data AnalystQuery writing and dashboard building are increasingly conversational.77%
  5. 05Web DesignerAI is turning standard website design into a low-cost commodity, while designers who solve difficult usability and business problems retain a stronger foothold; 2037 is a speculative displacement estimate, not a settled forecast.76%
  6. 06Web DeveloperStandard marketing and CRUD sites are generated in minutes.75%
  7. 07PHP DeveloperAI is rapidly commoditizing routine PHP development, making substantial role displacement plausible by 2036, though that date is speculative and developers who own architecture, security and production reliability remain harder to replace.74%
  8. 08Junior Software DeveloperRoutine implementation is heavily assisted; systems thinking and ownership still need people.73%
  9. 09System AdministratorInfrastructure is code and increasingly self-managing.70%
  10. 10Database AdministratorManaged services and autonomous tuning erode routine DBA work.68%
  11. 11Mid-Level Backend DeveloperAI is steadily absorbing routine API and database coding, but owning production reliability, security and messy integrations makes full replacement harder than generating working code.65%
  12. 12Senior Frontend DeveloperAI is rapidly commoditizing interface implementation, but replacing a senior frontend developer also means taking responsibility for architecture, accessibility and ambiguous product decisions; majority displacement by 2040 remains an uncertain forecast.64%
  13. 13Software EngineerAI is rapidly taking over routine coding, but replacing most software engineers also means mastering ambiguous requirements, fragile production systems and responsibility for failures—making 2040 a speculative forecast, not a deadline.64%
  14. 14UX DesignerWireframes and flows are generated easily; research and judgement are not.63%
  15. 15Senior Software EngineerAI is rapidly taking over implementation work, but replacing most senior engineers also means handing machines architectural judgment and production accountability; 2041 is a speculative scenario, not a reliable deadline.62%
  16. 16AI IntegratorAI integrators risk having their own tools automate routine integrations, making substantial displacement by 2039 plausible but uncertain while complex systems and accountable deployment remain more defensible.61%
  17. 17DevOps EngineerAI will increasingly write deployment pipelines and triage alerts, but owning production failures, security trade-offs and messy infrastructure migrations will keep experienced DevOps engineers valuable; majority displacement by 2040 remains an uncertain forecast.57%
  18. 18AI DeveloperAI can increasingly write the models and glue code, but developers who own real-world reliability, security and deployment remain harder to replace; majority displacement by 2040 is a speculative estimate.56%
  19. 19Product ManagerDocumentation and analysis compress, but taste, politics and prioritisation remain human.52%
  20. 20Cybersecurity AnalystAutomation handles triage, but adversaries also use AI — demand keeps rising.45%
  21. 21Robotics EngineerAI will write more robot code, but making machines work safely in messy real environments protects robotics engineers; majority displacement by 2050 is a speculative scenario, not a confident forecast.40%
  22. 22Technical LeadAI can write more of a technical lead’s code, but replacing the person who resolves architectural trade-offs and owns delivery is a much harder proposition; majority displacement by 2046 is a speculative estimate.39%
  23. 23Machine Learning EngineerThe people building the systems are among the last to be displaced by them.35%
  24. 24Chief Technology OfficerAI can shrink a CTO's research and reporting workload, but replacing the executive who owns technology bets, leads teams and answers for failures remains a much harder proposition; 2050 is a speculative endpoint, not a confident majority-displacement forecast.24%

Don't see your job?

Good to know

Will AI replace tech and software jobs?

Tech and software jobs share a particular exposure to AI: much of their work happens in digital systems where tools can generate, test and modify outputs. Across the 24 roles on this page, the average forecast risk is 62%, but the outlook varies substantially between routine execution and responsibility for complex systems. These forecasts are estimates, not deadlines for entire professions to disappear.

Why some tech jobs face greater exposure

The most exposed tasks have clear inputs, repeatable steps and outputs that are relatively easy to check. Manual software testing, basic support tickets, routine SQL reporting and standard website builds fit this pattern. AI coding assistants, automated test tools and support chatbots can take on parts of that workload.

This helps explain the higher forecasts for Manual Software Tester at 87% by 2029 and IT Support Technician at 81% by 2030. Exposure is not uniform within either role: reproducing an intermittent failure or untangling a customer's unusual setup requires more context than running a checklist.

What makes other roles more resilient

Less exposed roles tend to combine technical work with difficult trade-offs, organizational context or physical constraints. A technical lead negotiates architecture decisions across teams; a cybersecurity analyst investigates ambiguous activity; a robotics engineer must make software work reliably with hardware. AI can assist each, but checking recommendations and accepting responsibility remain substantial work.

The forecasts reflect that distinction: Chief Technology Officer is at 24% by 2050, while Machine Learning Engineer is at 35% by 2040. Neither title guarantees safety. Building dependable systems, evaluating failures and owning outcomes matter more than simply working on an AI product.

How to build a more durable tech career

Start by separating your workload into tasks AI can draft and decisions you must defend. Learn to review generated code, validate data, design meaningful tests and diagnose production failures rather than judging success by output volume. Use coding assistants without sending confidential code or customer information to unapproved services.

For testers, add test automation and exploratory testing. For support staff, deepen networking, identity and incident-response skills. For developers and analysts, build domain knowledge and take responsibility for deployment, data quality or stakeholder decisions. Juniors should seek code review and explain their reasoning, not merely deliver generated solutions.

Questions people ask

Will AI replace software engineers?
AI can automate parts of implementation, testing and documentation, but software engineering also involves requirements, architecture, debugging and operational responsibility. This page forecasts a 64% risk by 2040 for Software Engineer; that is not a claim that all software engineering jobs will disappear by then.
Is tech still a safe career?
Tech is not uniformly safe or uniformly threatened. Look for roles where you learn to verify AI output, solve unfamiliar problems and take responsibility for systems people depend on, rather than relying solely on routine digital production.
Which tech jobs are least likely to be replaced by AI?
Among the roles listed here, Chief Technology Officer, Machine Learning Engineer, Technical Lead and Robotics Engineer have the lowest forecast risks. Their relative resilience comes from leadership, complex evaluation, cross-team decisions or physical integration, not immunity to automation.
Should I still learn to code if AI can write code?
Yes, if you want to build or maintain software: understanding code is essential for spotting incorrect logic, security flaws and fragile assumptions in generated output. Pair programming fundamentals with debugging, testing, version control and experience deploying something real.