Could rogue AI agents evade Singapore's safety guardrails? | Deep Dive

AI agents can plan gym workouts, book flights and make decisions on our behalf. But granting them greater autonomy may pose serious consequences when AI agents go rogue. With AI researchers calling for tighter restrictions, what safety guardrails should Singapore put in place - and are we already doing that?

In this week's Deep Dive, Steven Chia and Tiffany Ang sit down with leading scholar on AI in financial economics Professor Will Cong from the Nanyang Technological University and Associate Professor Jennifer Ang from the Singapore University of Social Sciences, who sits on the AI Singapore Scientific Committee on Governance, to explore the risks of rogue AI agents and why human oversight remains essential even as AI is used to defend against AI threats.

Highlights:
00:00 Intro
03:13 How risky is everyday AI use?
07:20 What's new about us handing over our data?
09:30 Are humans still in control when dealing with AI agents?
11:08 Would you share your credit card details with an AI agent?
14:06 Why experts fear rapid AI adoption
16:17 How can we prevent AI from going rogue?
18:00 Can AI learn human values?
25:30 Is Singapore adopting AI too quickly?
27:46 Hard lines on what we should never let AI touch

===============
#DeepDivePodcast #AI #aiagents

Subscribe to our channel here: https://cna.asia/youtubesub

Visit our website: https://cna.asia
Download our app: https://cna.asia/app

Get our editors' picks of stories via chat apps:
Telegram: https://cna.asia/telegram
WhatsApp: https://cna.asia/whatsapp

Follow us on these platforms:
https://www.facebook.com/channelnewsasia
https://www.instagram.com/channelnewsasia
https://www.tiktok.com/@channelnewsasia