Securing the Invisible
summary
Shadow AI already causes measurable harms: data exfiltration that can trigger GDPR fines up to 4% of global revenue, rapid adoption (ChatGPT hits 1 million users in five days), and real incidents like the Okta support token compromise. Practical defenses include inventorying GenAI traffic via monitor DNS queries or web gateways (Zscaler/Cloudflare), enforcing approved tools with Defender for Cloud App + Defender for Endpoint, using Azure Content Safety or Mosaic guards, classifying data, and auditing vector stores (Pinecone) and model sources.
AI-written summary · about
chapters (7)
- 0:00Shadow AI risks
- 9:48Shadow AI dangers
- 16:15Data leaks and compliance
- 20:27Mitigating shadow AI
- 27:23AI security controls
- 33:05AI security platforms
- 39:14Shadow AI summary
related talks
34mInvisible Infiltration of AI Supply Chains: Protective Measures from Adversari...
42mSecure your AI application transformation with Security
49mThe Intersection of AI and Cybersecurity
52mDear CIO: Navigating the Shadows - GenAI's Promise, Peril, and the Path Forward
39mSecuring AI at Scale: Practical Defenses against Prompt...
21mThe Elephant in the Room: Turning Shadow AI into Corporate Value
49mGenAI and Cybersecurity: Safeguarding the Digital Age
39mThe Future of Cybersecurity in an AI-driven Era