The episodes have highlighted the vulnerabilities in AI security and raised questions over how the fast-growing technology can be developed safely as its usage becomes more widespread globally.