In a rapidly shifting labor market, a new study from Microsoft Research is sparking urgent conversations about how to build professional resilience in the AI era.
The paper, “Working with AI: Measuring the Occupational Implications of Generative AI,” identifies which jobs are most susceptible to disruption and which may offer more long-term security in the face of accelerating automation.
Released on July 10, the study analyzed over 200,000 anonymized conversations with Microsoft’s Bing Copilot. It introduces an “AI applicability score” to quantify how well AI systems can perform specific occupational tasks. Led by Senior Researcher Kiran Tomlinson, the team mapped the results against the U.S. Department of Labor’s O*NET database to evaluate the risk landscape across professions.
Interpreters and translators (score: 0.49), historians (0.48), and writers and authors (0.45) top the list of high-risk roles, all reliant on skills that generative AI is quickly mastering—namely language processing, summarization, and content creation. Customer service and sales roles are also flagged as vulnerable due to their reliance on repeatable communication patterns that AI can increasingly handle.
Meanwhile, jobs grounded in physical presence or manual labor—such as dredge operators, water treatment plant workers, and nursing assistants—show minimal overlap with AI capabilities, scoring below 0.03. These roles remain more resilient for now, reflecting AI’s current limitations in tactile, regulated, and highly interpersonal environments.
Still, Microsoft emphasizes that a high applicability score does not necessarily predict job elimination. “Our research shows that AI supports many tasks… but does not indicate it can fully perform any single occupation,” Tomlinson said. The paper frames AI as a productivity enhancer rather than a full replacement, comparing its likely impact to the way ATMs changed but did not eradicate bank teller roles.
The study’s release coincides with a pivotal moment for Microsoft. The company recently surpassed a $4 trillion market valuation, driven largely by AI investments, while laying off more than 15,000 employees in 2025 to realign around AI priorities—highlighting the tension between technological advancement and workforce stability.
Public figures weighed in quickly. Investor and commentator Codie Sanchez noted on X, “A few really surprised, and many we own,” in reference to the exposed job list. Others raised skepticism about whether AI should, or even could, replace entire professions. “Yeah, can’t wait for books written by LLMs trained on the internet,” wrote tech critic Kyle Beshears.
As job roles in research, writing, and communication become increasingly AI-integrated, the study urges workers and employers alike to focus on resilience: cultivating uniquely human skills, adapting workflows, and preparing for hybrid collaboration with AI tools. While there is only a weak correlation between wages and AI exposure, the broader message is clear—no role is entirely immune.
For those seeking to build resilience in the AI era, understanding where the technology overlaps—and where it falls short—is the first step toward navigating what comes next.
