Exploring the Intersection of AI and Security

Podcast: Play in new window | Download

Subscribe: Apple Podcasts | RSS

TechSpective Podcast Episode 132

The rapid adoption of generative AI technologies like ChatGPT has taken the tech world by storm. Generative AI models and the LLMs (large language models) they rely on are pervasive–and potentially invasive–which highlights the importance of AI safety research, emphasizing the need to accelerate efforts to keep pace with this swift adoption.

Adobe’s Strategic Approach

Gurpartap “GP” Sandhu, the leader of the product security AI and data engineering team at Adobe, joins this episode of the TechSpective Podcast to talk about how Adobe’s approach to AI is both cautious and innovative. GP delves into practical applications of AI within Adobe, such as auto-generating code fixes and scaling threat models using multimodal AI capabilities. These advancements not only improve efficiency but also enhance the overall security infrastructure. The discussion extends to broader societal impacts, including the challenges posed by deep fakes and AI-generated disinformation.

A Dual Mandate: Balancing Risks and Opportunities

GP emphasizes the concept of a dual mandate in AI and security. He explains that while it’s essential to focus on minimizing risks such as data poisoning and prompt injection, it’s equally important to leverage AI to enhance security measures. This balanced perspective sets the stage for a deep dive into how Adobe is navigating the complex landscape of AI in cybersecurity.

Encouraging Hands-on Engagement

We also talk about the importance of hands-on experimentation with AI technologies. They encourage listeners to engage directly with AI tools to better understand and contribute to their development. This call to action resonates with tech enthusiasts and professionals alike, emphasizing the collective effort required to navigate the evolving AI landscape.

Check out the full episode for more about AI security and security for AI. GP and I have some interesting back and forth about the potential risks and implications of AI, and finding a balance between being cautious and optimistic.

The podcast itself is audio-only, but the video of our conversation is also available on YouTube if you prefer:

Please ask questions and share your thoughts on the topic in the comments below. Also, please subscribe to the TechSpective Podcast through your favorite podcast platform and share the podcast with your peers and friends.

If you enjoy the podcast, I would also be grateful if you could take 2 minutes to rate and review the podcast on iTunes or wherever you listen.