AI in Healthcare: Potential Benefits Overshadowed by Structural Challenges, Researchers Warn

AI's Promise and Perils in Healthcare

Artificial Intelligence (AI) is rapidly making inroads into healthcare, promising revolutionary changes in various aspects of the industry. However, a recent study by researchers from the University of California, Santa Cruz and the University of British Columbia raises important questions about the technology's potential impact on public health 1

1

2

2

3

3

4

4

.

The Promise of AI in Healthcare

Proponents of AI envision numerous applications that could transform healthcare:

• Managing healthcare supply chains more efficiently
• Monitoring disease outbreaks in real-time
• Assisting with diagnoses and medical image interpretation
• Reducing equity gaps in healthcare access by compensating for worker shortages

AI could potentially improve resource allocation across hospitals, enhance patient triage, and expand the capabilities of general practitioners in rural areas lacking specialists 1

1

2

2

3

3

.

Concerns and Challenges

Despite these promising applications, the researchers, Lucia Vitale and Leah Shipton, highlight several concerns:

1. Privacy rights violations
2. Racial and gender biases in AI models
3. Lack of transparency in AI decision-making processes
4. Potential misuse by insurance companies for discrimination

The study argues that AI might become another technological advance with limited impact due to its engagement in a "politics of avoidance" 1

1

2

2

3

3

4

4

.

The "Politics of Avoidance"

Vitale and Shipton contend that AI in healthcare often focuses on treating diseases while ignoring underlying health determinants. This approach could divert attention from more fundamental structural problems in global public health 1

1

2

2

3

3

4

4

.

"We keep investing in these tech silver bullets that fail to actually change public health because they're not dealing with the deeply rooted political and social determinants of health," says Vitale 1

1

2

2

3

3

4

4

.

Exacerbating Existing Inequalities

The researchers warn that AI could perpetuate or worsen existing patterns of harm and exploitation in the healthcare industry:

• Concentration of AI ownership and profits in high-income countries
• Potential exploitation of low- and middle-income countries for data extraction or risky technology deployment
• Prioritization of intellectual property rights over equitable access to new treatments
• Neglect of the world's poorest populations in research and development 1

  1

  2

  2

  3

  3

  4

  4

The Need for Regulation

To maximize benefits and minimize harm, Vitale and Shipton argue for implementing regulations before AI further expands in the health sector. They call for:

• A clear agenda and framework for ethical governance of AI health technologies
• Involvement of the World Health Organization and major public-private partnerships
• Engagement of countries hosting tech companies, such as the US, India, and China
• Continued civil society advocacy 1

  1

  2

  2

  3

  3

  4

  4

As Shipton notes, "With AI, we have an opportunity to correct our way of governing new technologies" 1

1

2

2

3

3

4

4

. The challenge lies in ensuring that AI's potential in healthcare is realized while avoiding the pitfalls that have limited the impact of previous technological advances in public health.