AI and Cybersecurity
The discussion highlights the dual perspectives on AI in cybersecurity: using AI to enhance security measures and ensuring AI systems themselves are secure. Dinis emphasizes the current insecurity of many AI deployments, pointing out the lack of understanding around model behavior and potential vulnerabilities. He advocates for a cautious approach, suggesting that models should be read-only and not learn from new inputs to mitigate risks.In this clip
From this podcast
Practical AI
Cybersecurity in the GenAI age
Related Questions