Interviewed and published in the Mainichi Shimbun about the risks of conversational AI

Please see here for the detail.

● Mainichi Shimbun Online Edition (October 18, 2023)
[Challenging God - Where is human intelligence headed? "User malice can cause AI to behave in unintended ways" -Macnica 's Ryo Shota]
https://mainichi.jp/articles/20231014/k00/00m/040/113000c
(You must register to read the full article)

●Mainichi Shimbun electronic version (October 18, 2023)
[Human malice causing abnormal behavior that causes AI to “jailbreak”]
https://mainichi.jp/articles/20231014/k00/00m/040/102000c
(Registration required to read the full text)

Macnica has published an article explaining the techniques used by cyber attackers to abuse ChatGPT.