Australia’s Medical Association is calling for robust regulations and transparency regarding the use of artificial intelligence (AI) in the healthcare industry. This comes after doctors in Perth-based hospitals were warned against using ChatGPT to write medical notes. The Australian Medical Association highlighted the need for stronger rules to safeguard patients, healthcare professionals, and build trust. Five hospitals in Perth were advised to cease using ChatGPT after it was discovered that some staff members had been employing the language model for writing medical records. Patient confidentiality concerns were cited as the reason for discontinuing the practice. The AMA emphasized the importance of ensuring that clinicians have the final say in decisions influenced by AI, and that patients provide informed consent for any AI-based treatments or diagnostics. The association also stressed the need for secure protection of patient data and ethical oversight to prevent healthcare disparities.
The AMA suggested that Australia consider implementing the proposed EU Artificial Intelligence Act, which involves categorizing the risks associated with different AI technologies and establishing an oversight board. Additionally, they recommended exploring Canada’s requirement for human intervention points in decision-making. The AMA’s submission emphasized that human intervention must always be involved in important clinical decisions, ensuring that the final say rests with medical professionals rather than being a mere checkbox exercise.
AMA President Prof Steve Robson acknowledged that AI is a rapidly evolving field with varying levels of understanding. He emphasized the need to address the regulation gap, particularly in healthcare, where system errors, algorithmic biases, and patient privacy issues pose potential risks. Dr Karen DeSalvo, Google’s Chief Health Officer, expressed optimism about the potential of AI to enhance health outcomes but emphasized the importance of proper system constraints, factual accuracy, consistency, and adherence to ethical and equitable approaches.
A recent Google research study published in Nature revealed that their medical language model performed on par with clinicians, generating accurate answers to common medical queries 92.9% of the time. While AI has the potential to revolutionize healthcare, it is crucial to establish effective systems and practices.
In conclusion, the Australian Medical Association is urging the implementation of strong AI rules and transparency within the healthcare industry. They stress the need to protect patient welfare, ensure human involvement in critical decisions, and safeguard patient privacy and data. As AI continues to evolve, the focus should be on developing and implementing responsible regulations that enhance healthcare outcomes while addressing potential risks and challenges.