Hi Everyone,

Many companies today rely on traditional data pipelines that extract information from systems of record into data warehouses or lakes. While these pipelines centralize data, they often result in poor data quality—turning data into haystacks where finding operational information, let alone insights, feels like searching for needles.

To complicate matters, outdated data masking techniques are no longer sufficient against the power of generative AI (genAI), which can now re-identify masked datasets, posing serious security risks.

Confidential AI offers a solution. It provides a more secure and efficient way to manage sensitive data across multiple systems. With Confidential AI, you can enhance your existing data lakes and warehouses by unlocking data-in-motion and seamlessly connecting with various data silos. Best of all, data owners maintain full control by holding their own encryption keys, ensuring data sovereignty and privacy at every stage.

At Opaque Systems, we are proud to lead this transformation. Our Confidential AI platform enables you to securely access, share, and operationalize real-time data without sacrificing security or speed.

Simply Faster: Accelerate operations by securely processing data without additional steps like data masking, allowing for faster insights and cloud scalability.

Better Results: Use encrypted, high-fidelity data for more accurate AI outcomes, eliminating the need for masking or obfuscation.

Always Verifiable: Ensure privacy and compliance with cryptographic verification and a hardware root of trust, keeping your data secure and sovereign.

In this issue, we’ll explore how Confidential AI is reshaping data management and AI security, along with exciting updates from our lab.

— Aaron Fulkerson, CEO of Opaque Systems

Unlocking the Power of AI through Confidential Computing

In a world where technology constantly pushes the boundaries of what's possible, Larry Augustin, entrepreneur, angel investor, and former chairman of the board at SugarCRM, sees a future brimming with opportunity—one where confidential computing unlocks the full potential of AI to drive meaningful change across industries. 

By embedding confidential computing into our technological fabric, we can secure sensitive data during processing and open the door to innovative AI applications that were once held back by security fears, Augustin explained in an interview at Opaque’s Confidential Computing Summit. This shift not only promises to safeguard our digital world but also to empower organizations to harness AI in ways that truly make a difference.

Just as encryption standards like HTTPS and secure DNS have become ubiquitous, he predicted that confidential computing will soon be a standard practice across industries. This shift is driven by an increasing need to secure not just data at rest but also data in use, ensuring that sensitive information remains protected even during processing. This capability is crucial in enabling new applications that are currently hindered by security concerns. 

In financial services, the ability to share data securely across borders opens up new avenues for combating fraud and enhancing regulatory compliance, for example. In his interview, Augustin highlights how financial institutions can leverage confidential computing to collaborate on anti-fraud initiatives without risking data breaches or violating data sovereignty laws.

Early adopters, particularly in regulated industries like finance, will be critical for increasing adoption. These sectors are often at the forefront of security breaches, making them more likely to adopt new technologies like confidential AI. By doing so, they not only protect their data but also set the stage for broader industry adoption.

The integration of AI into everyday business processes is no longer a distant vision but a reality that is rapidly taking shape. However, for AI to reach its full potential, the security concerns surrounding data access and privacy must be addressed. Confidential computing offers a robust solution, enabling the secure execution of AI applications and paving the way for new, actionable use cases.

Watch the interview with Larry Augustin below.

In the Lab

The Latest Happenings at Opaque Systems

Whitepaper: Solving Digital Transformation’s Data Security and Privacy Problems

By the numbers: The average enterprise collects data from 400 different sources, and its data holdings are growing at 63% per month. To future-proof data operations and uncover maximal business value, enterprises need to deploy a confidential and secure digital infrastructure like a confidential AI platform to enable secure sharing and analysis of data among internal teams, across business units, and with external partners. To learn more about this solution, download our latest whitepaper. 

TechStar Finale Featuring Raluca Ada Popa

Raluca Ada Popa, associate professor at UC Berkeley and co-founder of Opaque Systems, delivered an inspiring keynote address at Accenture's TechStar virtual graduation, following a warm introduction by Accenture CEO, Julie Sweet. Speaking to an audience of over 2,000 attendees, Raluca shared insights from her personal journey, discussing the challenges she faced along the way and the valuable lessons she learned. She also highlighted the transformative potential of generative AI in driving innovation, as well as the crucial role confidential computing plays in ensuring data security while equipping leaders with the insights needed to shape the future of work.

5 Ways to DORA-Proof Your Business with Confidential AI

DORA, set to take effect in early 2025, aims to fortify the EU’s financial sector against the growing threats of cyberattacks and data breaches. Opaque’s user-friendly solution provides the tools necessary to not only comply with DORA’s requirements, but also to enhance digital resilience, and make it quick and easy to do so. Read our recent blog post to learn how businesses can ensure they are fully DORA-compliant by adopting confidential AI.

Solution Brief: Navigating the Complexities of AI, Analytics, and ML

Thanks to confidential AI, previously inaccessible data, trapped in silos, bound by regulations, or limited by corporate requirements, can now be put to work across industries. Discover how the functions and features of Opaque Workspace—the leading confidential AI platform—work to speed up project timelines and accelerate AI projects into production.

Career Snapshot: Life Lessons in Leadership and Innovation

Aaron Fulkerson, part of the inaugural cohort of the Cooke Transfer Scholars at the University of North Carolina at Chapel Hill, shares his journey from a first-generation college graduate to accelerating the future of AI innovation as the CEO of Opaque Systems. 

Stay Ahead of the Curve: Industry Solutions

Last month, we updated the Opaque Systems website with a new Industry Solutions resource. Tailored specifically for the high-tech, financial services, and insurance sectors, these pages uncover how organizations can leverage cutting-edge confidential AI innovations to drive competitive advantage, ensure data privacy, and stay ahead in the rapidly evolving tech landscape.

The Enterprise AI Sourcebook is Here
We haven’t even scratched the surface regarding reaching AI’s potential. If you have a creative mind, the willingness to learn, and the ethical imperative, the possibilities are endless. The Enterprise AI Sourcebook features articles by experts, discussing developments within their fields, including insight on prompt augmentation and retrieval-augmented generation from our CEO Aaron Fulkerson.

Code for Thought

Worthwhile Reads

⚖️ The upcoming enactment of California’s AI regulation bill marks a pivotal moment for the industry. Representing one of the first comprehensive attempts by a major jurisdiction to regulate AI technologies, SB 1047 will require AI developers to comply with certain rules before developing their models, setting new standards for transparency, accountability, and ethical use of AI technologies. The bill continues to divide big tech: critics like OpenAI argue that the federal government, not individual states, is best suited to regulate AI, while backers of the proposal note that Congress has thus far failed to act. 

🚗 FTC targets data privacy, taking action against companies that use connected car data. Automakers can track everything—from driving patterns to music preferences—through extensive data collection. The lack of strong legal protections makes it a fight to safeguard data—spotlighting the increasing significance of confidential computing as a viable solution to protect sensitive information. A recent data-collection lawsuit against General Motors Co. could signal increased regulatory scrutiny of connected vehicles in the future. 

🧠 Scientists are beginning to unravel the "black box" of genAI models. As these models are used in critical applications, it is crucial to comprehend their inner workings and how they produce specific outputs. One way AI researchers are trying to understand how models work is by looking at the combinations of artificial neurons that are activated in an AI model's neural network when a user enters an input. This research aims to enhance AI safety and ensure reliable decision-making by making model behavior more transparent—confidential AI can work into this research, ensuring that this process occurs without compromising data privacy or security. 

🔐 Meta has created a new suite of security benchmarks for LLMs. Designed to benchmark AI models’ cybersecurity risks and capabilities, the CyberSecEval-3 framework outlines five key strategies. “Compared to previous work, we add new areas focused on offensive security capabilities: automated social engineering, scaling manual offensive cyber operations, and autonomous offensive cyber operations,” Meta researchers wrote.