Controlling Rogue AI: Rakuten’s Approach to AI Security
AI models are designed to mimic human behavior and manage complex tasks, but AI can become rogue when they disregard commands, share misinformation, use menacing language, or conduct cyberattacks. Rogue AI can occur when hackers compromise the AI model's confidentiality, integrity, or availability, and this session will highlight how Rakuten is addressing these risks by balancing innovation with secure and ethical measures.【Time】24:53
Chapter 1
Self-Introduction
Chapter 2
Introduction to Cyber Security Defense Department
Chapter 3
Rouge AI and the Definition of AI Security
Chapter 4
Documents for AI Security
Chapter 5
AI Red Teaming
Commerce & Marketing Company
Group Company Division (Incubation Business)
Technology Divisions
19:50
Identity Management and AI
24:53
Controlling Rogue AI: Rakuten’s Approach to AI Security
32:28
Evolving Product Management with AI and Data
20:28
Sculpting Horizons: The Art of Front-end Governance & Platform Engineering
30:40
A Technical Deep Dive into Rakuten's Data Ocean
32:00
The Computer Vision R&D Reinforcing Rakuten Group's Future
32:18
From Our Ireland Branch: Rakuten Blockchain Case Study and Web3 Undertakings
