AI in healthcare needs a rethink: Doctors struggle with its burden

Artificial intelligence (AI) is rapidly entering primary care, promising to enhance diagnostics, streamline workflows, and reduce the administrative load on healthcare providers. However, a new study suggests the reality is more complex. The research, titled “Artificial Intelligence in Primary Care: Support or Additional Burden on Physicians’ Healthcare Work?—A Qualitative Study” and published in the journal Clinics and Practice, provides critical insights into how AI impacts physicians’ daily routines, decision-making, and patient relationships.

Through in-depth interviews with 28 primary care physicians across Germany, the researchers examined whether AI tools function as helpful support systems or create additional pressures in clinical settings. Their findings reveal both significant potential and persistent challenges.

How do physicians perceive AI integration in primary care?

The study reveals a clear ambivalence among primary care physicians. On one hand, AI is recognized as a valuable aid in diagnostics, preventive care, and administrative tasks. Physicians reported that AI-powered tools can speed up the identification of conditions, support evidence-based decisions, and reduce time spent on repetitive documentation. These advantages are particularly meaningful in rural or underserved areas, where resource constraints limit access to specialist expertise.

Yet, many physicians remain cautious. They worry about the opacity of AI systems, called “black boxes”, and their potential to undermine clinical autonomy. When algorithms deliver recommendations without transparent reasoning, physicians find themselves balancing trust in the tool against their own expertise. This tension leads to increased cognitive load, particularly when decisions carry high stakes.

The perception of AI as either a helpful assistant or an intrusive supervisor depends largely on how well the technology integrates into existing workflows. Poorly designed systems can create more work than they save, forcing physicians to navigate irrelevant alerts and disruptive interfaces.

What are the risks and challenges of AI use in healthcare?

While AI promises efficiency, the study highlights several risks that can turn these systems into burdens. Alert fatigue emerged as a major concern, with physicians reporting that constant, often unnecessary notifications erode concentration and create stress. Instead of improving productivity, such alerts may slow down decision-making and diminish trust in the technology.

Ethical and legal concerns also loom large. Physicians fear being held accountable for errors made by AI systems while lacking the transparency needed to justify decisions. Data security and algorithmic bias are further sources of anxiety, as the misuse or misinterpretation of patient data could exacerbate health inequities.

The research points to a psychological dimension as well. Insufficient training in AI use leaves many physicians feeling unprepared and anxious about integrating these tools into their practice. This “technostress” is compounded by the fear that overreliance on AI might erode the doctor-patient relationship. Instead of enhancing care, some worry AI could depersonalize interactions, making consultations more mechanical and less human-centered.

These concerns highlight the need for explainable, user-friendly AI systems and clear guidelines on accountability and ethical use.

What conditions are needed for AI to truly support physicians?

The study underscores that AI’s impact is not predetermined, it depends on how the technology is implemented. Physicians emphasized several conditions for successful integration. First, AI systems must be transparent, with algorithms that provide understandable reasoning behind their recommendations. This would build trust and allow clinicians to use AI as a supportive second opinion rather than an opaque directive.

Second, user-friendly design is essential. Tools should complement, not disrupt, existing workflows. When AI integrates seamlessly into clinical practice, it can automate routine tasks and free up time for meaningful patient interactions. Conversely, clunky systems risk increasing workload and frustration.

Third, training plays a critical role. The authors recommend embedding AI literacy into medical education and providing ongoing support for practicing physicians. This would empower clinicians to use AI confidently, understand its limitations, and maintain ultimate control over clinical decisions.

Finally, successful adoption requires involving physicians in the design and implementation of AI tools. When users have a voice in shaping the technology, systems are more likely to meet the real needs of primary care environments.

Overall, the study asserts that for AI to truly serve as a support system rather than an added burden, developers, policymakers, and healthcare institutions must prioritize transparency, usability, and clinician involvement. Ethical guidelines and robust training programs will be essential to ensure that AI strengthens rather than undermines the humanistic core of primary care.