AI Resilience Report forData Scientists

Many data science tasks are being helped by AI, but the core job isn’t vanishing. For example, data scientists spend 60–80% of their time cleaning and preparing data, and new AutoML tools can now automate many of those steps ^[1]. Official sources say data scientists “use machine learning to extract and analyze information from large … datasets,” then “visualize, interpret, and report data findings” ^[2].

In practice, this means AI can assist with coding models or handling routine data queries – Microsoft’s research even lists “Data Scientists” among jobs where generative AI can perform many tasks ^[3]. In short, simple analysis and coding are easier now because of AI. But the hardest parts – like deciding what questions to ask, understanding which data matter, and explaining insights – still need people.

McKinsey notes that the first steps of any data project (framing the business problem and choosing data) require human insight ^[1]. In other words, AI is automating some parts of a data scientist’s work, but human skills in judgement and storytelling remain essential.

AI tools for data work are widely available (cloud platforms and open-source libraries), so companies can start using them without huge software budgets. Because data scientists are paid very well (median ~$112,600 ^[4]) and in short supply ^[1], firms have a big incentive to boost productivity with AI. For instance, McKinsey reports that data-scientist job postings have more than tripled since 2013 ^[1].

At the same time, adopting new AI systems involves costs (buying computing power, ensuring data privacy, training staff). Businesses also care about trust and ethics. Notably, a Microsoft study stresses that high AI use tends to change how work is done – not necessarily replace jobs ^[3].

Companies often see AI as an assistant for experts rather than a substitute. In fact, BLS projects very strong growth (34%) in data science jobs ^[4], suggesting human demand remains high. Overall, data teams are likely to adopt AI tools gradually: they’ll use AI to handle repetitive analysis and speed up work, but still rely on skilled people for planning, checking results, and communicating findings in context.