Most healthcare workers don't check security protocols before trying out new generative AI tools such as ChatGPT, putting patient and other sensitive data at risk, said Sean Kennedy of software vendor Salesforce, which recently conducted research on potential security gaps in healthcare settings.