Over the past 3 or so years, healthcare advancement has been at the forefront of the media, with fast work developing vaccines and hard work reducing hospital times. AI has come to revolutionise this industry, along with practically every other. The benefits of new technologies within medical organisations today are becoming more and more evident, with even the government releasing an article about AI at the front of the COVID-19 crisis.
AI in healthcare may be revolutionary, but it does also pose a lot of threats. Cybersecurity is a crucial part of the healthcare industry with over half of healthcare organisations worldwide suffering from ransomware attacks in 2021. AI is so advantageous to these companies that there is a new thin line between security and progression.
The accelerated progression that AI brings to healthcare is supreme. AI can understand genetics and medical history to determine what a patient may be predisposed to: “Deep learning models can analyse scans and X-rays to diagnose patients with 90% accuracy“. This speeds up wait times and hospital trips significantly, whilst helping the doctors gain a thorough understanding of the diagnosis.
AI is also being used to test how chemicals might interact, this helps to develop new medicine by running experiments quickly and cost-effectively. These discoveries can now take less than a day; an article about AI in healthcare describes the supercomputers that predict effective medicines through their database of molecular structures and disease.
AI also helps with administration, including data entry and organisation tasks. These systems are now “helping consolidate records and make them more accessible while saving time for hospital staff. Doctors and nurses can then find relevant information faster and spend more time helping patients.”
These are all very clear and note-worthy advantages, but there is also the consequence of increased risk. The data required for all of these tasks that the AI systems can undertake in minutes is vast. This data is bound to be attractive to cybercriminals and faces a large risk of breaches.
Nurses and Doctors need quick and easy access to files and resources – however, secure data will, in the meantime, be challenging to access quickly.
Medical record information is private and sensitive with no exceptions. Because of this, companies will face huge legal repercussions if there is a leak in data. The more that AI is used to generate solutions, the more it becomes likely hackers may come and “poison” the data to produce false readings and results, leading to medical errors which threaten people’s survival.
There are a few ways Healthcare can Secure the future of AI, these include; Better Data Stewardship, Stricter Access Controls, Employee Training, Monitoring and Security.
Data Stewardship is “the management and oversight of an organization’s data assets to help provide business users with high-quality data that is easily accessible in a consistent manner.” By securing the data in the algorithms used for AI within the healthcare sector, we will begin to strengthen the system as a whole. There are two ways of De-identification in The US Health Insurance Portability and Accountability Act (HIPAA): Expert Determination and Safe Harbor.
Using either of these in an AI algorithm will help to ensure errors or breaches don’t expose sensitive patient data. Organisations should regularly clean their data and encrypt databases to reduce the risk of hacking.
Access must be restricted, this helps to reduce data leaking. The best way to tackle this is to follow the Least Privilege Principle. Privilege should only be given to those who need it to complete a task. If a person doesn’t need access to the information they should not be given the right to have access to it. This is otherwise seen as a “need to know” rule that helps to reduce the number of paths that lead to sensitive information. To keep these platforms and data secure, businesses must regularly change their passwords and implement Multifactor Authentication.
Human error produces the majority of data breaches in the healthcare sector, the best way to get around these mistakes is to train employees on how best to use the system, to avoid data breaches. There should be regular cyber security training and an understanding of the importance of this practice among staff members. The training should include demonstrating the importance of password management and explaining to staff the privilege access theory.
Monitoring should be put into place to look for suspicious activity. If the systems are being monitored there will be a faster response rate to an attack. AI can be used for this to provide 24-hour support that would be otherwise difficult for IT staff to work. AI can detect potential breaches a lot faster and contain them before there is any significant damage.
The wealth of information that AI can store and produce for us every day is incredible. When used in the medical sector, it has such great powers and an ability to help to develop support and healthcare solutions for patients and staff. The risks and challenges that come with storing the data from these aids are a challenge that as a society we haven’t faced in such extremes ever before. “Every two days now we create as much information as we did from the dawn of civilization up until 2003” So we are bound to face challenges when it comes to storing and ensuring cyber security on that data. It is important that we begin to understand the most sensitive ways of accessing data and protecting it before it unintentionally gets into the wrong hands.
“Deep learning models can analyse scans and X-rays to diagnose patients with 90% accuracy”
