AI tools are now baseline

Employers are increasingly treating artificial intelligence coding tools as standard equipment for engineers, not an optional extra. The clearest signal is in developer surveys: 84% of respondents in Stack Overflow’s 2025 survey said they use or plan to use artificial intelligence tools in development, and 51% of professional developers said they use them daily. (survey.stackoverflow.co) GitHub found a similar pattern inside large companies. In its 2024 enterprise survey of 2,000 people on software teams across the United States, Brazil, Germany, and India, more than 97% said they had used artificial intelligence coding tools at work at some point, and 88% of United States respondents reported at least some company support for using them. (github.blog) That helps explain why hiring chatter now focuses less on whether engineers use these systems and more on which systems they can work with. Cursor, one of the best-known code editors built around artificial intelligence assistance, now sells a $20-a-month Pro plan and a $40-per-user Teams plan with shared rules, usage analytics, privacy controls, and single sign-on for employers. (cursor.com) The same shift is spreading beyond software engineering into data work. Apache Airflow, the open-source workflow tool widely used by data engineers, released Airflow 3.0 on April 22, 2025, and the project’s 2025 survey drew 5,818 responses from 122 countries, underscoring how large the installed base is for teams now layering machine learning and large language model workflows onto existing data pipelines. (airflow.apache.org) One example came from Astronomer, the main commercial company behind Airflow, where a support engineering lead wrote in June 2025 that his team used Airflow 3.0 and the Airflow artificial intelligence software development kit to build a large language model system that helps engineers investigate customer tickets faster. He said the goal was to speed up expert work, not replace support engineers. (medium.com) Hiring demand is also tilting toward workers with explicit artificial intelligence skills. LinkedIn data published through the Organisation for Economic Co-operation and Development showed that 7 out of every 1,000 LinkedIn members globally counted as artificial intelligence engineering talent in 2024, up 130% from 2016, while artificial intelligence literacy skills such as ChatGPT and Microsoft Copilot grew 600% in the last year. (oecd.ai) By January 2026, LinkedIn said artificial intelligence had added 1.3 million new roles globally, including more than 600,000 artificial intelligence-enabled data center jobs, even as overall hiring remained about 20% below pre-pandemic levels. The same report said artificial intelligence engineer was one of LinkedIn’s fastest-growing job titles over the prior three years. (weforum.org) Public labor data points in the same direction for technical roles. The United States Bureau of Labor Statistics said on March 11, 2025 that artificial intelligence is likely to augment many computer jobs, and it projected software developer employment to grow 17.9% from 2023 to 2033, compared with 4.0% for all occupations. (bls.gov) Employers are not reading that as a reason to trust the tools blindly. Stack Overflow’s 2025 survey found that only 33% of developers trust artificial intelligence output, while 46% distrust it, and just 3% said they highly trust it. (survey.stackoverflow.co) So the new baseline is not “let the model code everything.” It is closer to “use the tools, move faster, and still check the work” — the kind of force-multiplier expectation that is now showing up in product pricing, enterprise rollouts, data engineering stacks, and labor-market data at the same time. (survey.stackoverflow.co; github.blog; bls.gov)