2553: Fortanix: Why Confidential Computing Will Be the Underlining of Training Data for AI

In the rapidly evolving landscape of artificial intelligence and data science, today's episode of the Tech Talks Daily Podcast brings forth a critically important yet often overlooked aspect—securing data through its entire lifecycle. Our esteemed guest is Anand Kashyap, the CEO and Co-founder of Fortanix, a company that has pioneered in the realm of confidential computing.

Anand and his team at Fortanix have uniquely positioned themselves in the industry, focusing not just on securing stored or in-transit data, but also data while it's being processed.

As AI models increasingly become the linchpins for competitive advantage, it's no longer just about the intricacies of these models but the treasure trove of data that fuels them. Giants like Google and Microsoft, armed with vast amounts of data, seem to have an edge. Yet, this edge brings its own challenges around the ethics, privacy, and compliance of using such data—questions that are resonating loudly, not just in the EU under the shadows of regulations like GDPR, but across the globe, including the US.

During our conversation, Anand demystified the concept of confidential computing, a term that is gradually becoming a cornerstone in the discussion around data security. At its core, confidential computing involves leveraging hardware-based trusted execution environments to protect both data and code. One might wonder how this impacts the realm of privacy regulations. Interestingly, confidential computing provides an effective, transparent mechanism to demonstrate compliance without diminishing the operational integrity of AI models.

As the conversation unfolded, Anand offered valuable insights, pointing out that while confidential computing does augment runtime security, it is not a silver bullet. The necessity for well-crafted applications and comprehensive security measures remains unabated. He emphasized that this form of computing, though robust in security mechanisms, does not negate the need for diligent application development and maintenance.
We also touched upon the far-reaching implications of data privacy laws like GDPR on the global architecture of data solutions.

Anand highlighted how such regulations are catalyzing the adoption of confidential computing solutions across different jurisdictions, with an intriguing possibility of creating sovereign cloud zones within public clouds.

Ethical considerations in data protection strategies occupied a significant segment of our discussion. With a discerning perspective, Anand shared his thoughts on deploying confidential computing for a myriad of new use cases, from healthcare to financial services. Anand, who continually evolves through self-education, including listening to thought-provoking podcasts, left us with an enriching viewpoint on the challenges and opportunities lying ahead in the data security spectrum.

As the spotlight continues to shine on AI and data, understanding how to secure this data effectively is not just an operational need but an ethical imperative. This conversation with Anand Kashyap illuminates that pathway, shedding light on both the potential and the limitations of confidential computing.

Whether you are a business leader, a tech enthusiast, or someone deeply involved in data privacy concerns, this episode offers a comprehensive lens through which to view the evolving narrative of data security.