The Pros and Cons of AI in the Health Sector: Data Privacy Is at Risk

There is no doubt that AI has completely revolutionized the health sector. Statistics show that its adoption can save up to 10% of costs and by 2030, the market for AI in healthcare is projected to reach $187.95 billion. AI has many benefits, but detractors are consistently cropping up. They warn of big issues like data breaches and false hype. For instance, in a recent paper published in the journal Lancet Oncology, nine leading cancer doctors and academics said that despite the hype of AI, public health systems are at a tipping point in cancer care, with AI being incapable of addressing the fundamental issue of cancer as a systems problem. AI is a tool, they say, but one whose efficiency for boosting actual survival rates is far from proven. This groundbreaking technology has another downside: data breaches, which can increase the inequalities in health care and add to the already unbearable burden experienced by the sector.

The Benefits of AI

AI has been advancing in a host of healthcare fields and several AI technologies have been harnessed on a mass scale. This technology is used to assist diagnostics and clinical decision-making in numerous specialities, including dermatology, pathology, genetics, robotic surgery, ophthalmology, and transplant surgery. AI can be employed to detect numerous problems and help physicians with decision-making. 

A Big Need for Data

In order to be used for these tasks, AI requires a large amount of data. Protocols for data encryption and sharing for AI-based research are currently decentralized. In other words, protocols are decided on a case-by-case basis and approved or disapproved by specific institutional ethics committees. For instance, for an individual AI-based study using patient data retrospectively, an ethics committee may decide to waive the need for patient consent. This has led patients to question AI’s ability to protect data. One recent survey has shown, for instance, that American patients are divided as to whether they are comfortable with AI having access to their data. Despite 100% of patients now having AI involved in their healthcare experience, only 38% of them say they trust this technology. Privacy isn’t the only problem. Bias, inequality, and the misuse of AI-driven tools are also causing great concern.

The Downside of AI

AI can be used by cyber criminals in various ways, including phishing emails and AI-created malware code. In the most recent report crafted by the Department of Health and Human Services (HHS) for Congress, evidence showed that in a given year, the HHS Office for Civil Rights receives over 600 notifications of data breaches that affect 500 or more individuals. What’s more, there are over 63,000 reports about data breaches affecting fewer than 500 people. The vast majority of breaches are caused by phishing emails. AI can be misused to create perfect, unsuspicious phishing emails, as they are written in perfect grammar, without the typical grammatical and spelling errors that could easily indicate a phishing ploy. It can also be utilized to gather massive amounts of stolen data from numerous sources and use personal information to create customized phishing e-mails. 

Proactive Measures to Take

FBI Director, Christopher Wray, has warned the public that cybercriminals are weaponizing AI and that attacks are predicted to become more sophisticated and difficult to overcome. To reduce data breaches, various steps need to be taken, including data minimization (the collection of data that is absolutely necessary for the intended purpose), encryption, the implementation of robust access controls to restrict access, anonymization of data before it is shared, the establishment of clear data sharing agreements, and the definition of data retention periods. Equally necessary are the implementation of robust security measures (including firewalls and monitoring mechanisms), compliance with privacy regulations, the development of comprehensive data breach response plans, and carrying out regular security audits and assessments.

AI is already deeply ingrained into healthcare systems across the globe. Numerous breaches have raised a red flag for how this potentially lifesaving technology is used. Healthcare systems must adopt a multifaceted approach to quell data breaches, the majority of which are caused by phishing. Strategies to embrace include minimization, access restriction, and the use of regular security audits as a means of prevention.