AI Use Quietly Erodes Professionals' Skills Through Insidious Deskilling, Warns New Research

AI Reliance Creates Hidden Professional Skill Erosion

Artificial intelligence has become an indispensable assistant across professional fields, with doctors, researchers, and software developers increasingly integrating AI tools into their daily workflows1

1

. However, new research published in the journal Human Affairs reveals a troubling consequence: AI use is causing deskilling, where professionals' skills deteriorate as they outsource critical tasks to artificial intelligence systems2

2

. Kaisa Kärki, an Academy Research Fellow in theoretical philosophy at the University of Helsinki, warns that this phenomenon operates insidiously, making it difficult for professionals to notice the undermining of professional skills in their everyday work1

1

.

Overreliance on AI Stems From Cognitive Biases

The research identifies a core problem: biases in AI reliance lead professionals toward overreliance on AI systems. When large language models respond convincingly, users develop unwarranted trust despite these systems making mistakes2

2

. Kärki illustrates this with healthcare scenarios where a doctor relying on a large language model without catching its inaccuracies could commit medical errors. This raises thorny questions about responsibility. Traditional philosophical theories suggest agents bear responsibility for actions taken with awareness, but the inconspicuous nature of deskilling challenges these frameworks1

1

. Professionals often fail to notice their own biases or excessive dependence on AI, making accountability unclear when skill decay leads to mistakes.

Deskilling Represents a Societal and Structural Problem

Kärki emphasizes that because deskilling easily goes unnoticed, individual professionals cannot shoulder all responsibility for related problems. Instead, deskilling represents a societal and structural problem requiring collective action2

2

. Professional communities must identify core skills that should remain in human hands. Aviation provides a model: pilots regularly practice responding to autopilot failures. Other professions should ask similar questions about functioning without AI when necessary1

1

. In medical education, experts already advocate for AI-free stages in professional education where students develop fundamental abilities. Diagnostic reasoning exemplifies a skill central to medical practice that requires human mastery, along with general field understanding2

2

.

Urgent Action Needed to Prevent Irreversible Skill Loss

The research, supported by the Weisell Foundation, University of Helsinki, and Research Council of Finland, warns that unchecked AI adoption at workplaces could leave society incapable of training new professionals without AI assistance2

2

. This dependency risks overlooking AI errors and threatens community resilience, which depends on diverse human skills. Kärki's article titled "Who is Responsible for Tackling Professional Deskilling?" argues that professions must preemptively address skill decay before it reaches community-level crisis1

1

.

Comprehensive Solutions to Combat Professional Deskilling

The research outlines actionable strategies to preserve professionals' skills. Awareness campaigns must educate workers that AI use can insidiously cause deskilling. Even experienced professionals should undergo regular skill testing to identify deterioration early2

2

. Organizations should designate certain tasks as human-only, prohibiting AI outsourcing entirely. Workplaces need AI use guidelines explicitly addressing skill decay risks. Higher education programs should incorporate mandatory AI-free training stages where students develop foundational competencies without technological assistance1

1

. Financial incentives could reward skill maintenance, while legislation should regulate AI use in fields where deskilling threatens critical societal processes. Professional communities must determine which core skills require human preservation. Finally, researchers should expand studies of skill decay beyond medicine to understand how AI affects software developers, researchers, and other knowledge workers2

2

. As AI integration accelerates, these preventive measures become increasingly urgent to maintain professional competence and societal resilience.