Hello!

Last year I spoke to an inspiring group of confidential computing innovators, users, solution providers, technologists, and academics from the stage at the first annual Confidential Computing Summit in San Francisco. The event showed me—in the amount of registrations alone—that we were on to something. And the market numbers only reinforce that fact: according to some estimates, the confidential computing industry will balloon from $5.3 billion in 2023 to nearly $60 billion by 2028.

Since I spoke last year, GenAI has improved, data sets have grown, and bad actors have been able to leverage AI itself to bypass certain data protection measures. A majority of enterprises have either adopted or hope to adopt an AI component to their offerings, yet many are stuck between the desire to innovate and the need (regulatory or otherwise) to keep data secure. I recently wrote an ACM Queue article comparing confidential computing to cryptographic computing and found that the best practical data security measure for AI remains platforms that run on confidential computing hardware.

I am immensely proud of the work we are doing at Opaque Systems to bring this type of platform to life: proving that innovation and security do not have to be at odds. I hope to see you from the stage next week where we will again unpack the ever-growing need for confidential data and AI solutions.

— Raluca Ada Popa, co-founder and president, Opaque Systems

Discounted Registration at Confidential Computing Summit

Attend the conference on June 5-6 in San Francisco to learn about the latest advances in GenAI security and safety from high-profile speakers like Azure CTO Mark Russinovich, Anthropic CISCO Jason Clinton, NVIDIA CSO Dave Reber, Group CEO of Accenture Kathik Narain, UC Berkeley academics, and many others. Please enjoy a 65% discounted registration here!

In the Lab

The Latest Happenings at Opaque Systems

Opaque Systems at NYSE Tech Summit

Opaque Systems was front and center at the NYSE Tech Summit in San Francisco. Ryan Hebert, BISO, NYSE and FIDS shared on a panel about why Opaque is high on his radar. And co-founder Ion Stoica—the man behind industry-changing frameworks Spark and Ray–took the stage to chat through the interplay of data security amidst rapidly emerging technology. Stoica dove deep into the ethical implications of regulating AI along with the evolving need for technical understanding of the landscape as a whole. Watch the clips here:

Raluca Ada Popa Teaches First-of-its-kind Course for Accenture

Raluca Ada Popa, Opaque Systems co-founder and president, and professor at UC Berkeley, recently co-taught the first generation of a course for executives covering GenAI systems and the security of large language models. Along with her colleagues at Berkeley SkyLab, Popa taught the course to Accenture executives, dubbed the Accenture-UC Berkeley Foundation Model Architect Program. The course, which will eventually be freely available to the public, explored topics around GenAI trustworthiness including security, privacy, and safety.

Product Demo: Privacy Preserving Cross Border Detection

See Opaque System’s product in action with a demonstration of how the platform can share malware data to detect common attacks without directly sharing individual malware data sets. We cal this our “share don’t show” use case. See a live demo here.

CEO Aaron Fulkerson at The Artificially Intelligent Enterprise Conference

Opaque Systems CEO Aaron Fulkerson spoke a few weeks ago at TechStrong Events virtual conference “The Artificially Intelligent Enterprise.” The event featured top industry experts discussing the latest in AI and the enterprise. Fulkerson’s talk focused on accelerating AI projects from pilot to production while stressing the necessity for ensuring data privacy, security, and sovereignty. Register to watch the session on-demand here.

Code for Thought

Worthwhile Reads

🦺 OpenAI forms safety group to oversee security decisions: A new committee will advise OpenAI’s full board on crucial safety and security decisions for upcoming projects. The organization on Tuesday announced the committee as it begins training a new AI model that will replace the GPT-4 system. The committee comes after an OpenAI researcher resigned and accused the organization of deprioritizing security measures in favor of flashy AI products. 

📈 Data breaches rose to over 78% in 2023: New research spotlights the impact of cyber threats and reinforces the need for improved data security measures across industries. The Identity Theft Resource Center released a report analyzing data breach trends in 2023 and found that leaks and exposures hit 3,205 and impacted more than 353 million people. This marks a 78% year-over-year increase from 2022 and a 72% increase from the record set in 2021. 

📑 Vermont’s new privacy law allows people to sue companies: Vermont has passed one of the nation’s most robust data privacy bills, which allows consumers to sue companies for violating their privacy rights. The law, which passed in May, severely limits the personal data businesses can collect and use. It also prevents companies from selling consumers’ sensitive data and permits people to sue companies they believe have overstepped without relying on the state to take action. 

⛓ Corporations cite data security as the key challenge of AI usage: Data privacy is the main concern for companies when asked about their current AI sentiments. New research conducted by AI analytics platform Altyrex finds that 80% of companies cite data security and privacy as the primary risk of scaling AI at their organizations. Additionally, nearly half of the companies surveyed shared they experienced inadvertent data exposure while deploying new AI solutions. 

🤷‍♂️ New research finds most CIOs not fully onboard with GenAI deployment: A new study conducted by professional services network EY finds that a majority of CIOs acknowledge the potential and impact of GenAI for their business, but aren’t fully convinced of its immediate value. The organization’s 2024 EY CIO Sentiment Survey found that only 26% of CIOs reported they deployed AI in a meaningful way, while nearly half are in the experimentation or proof-of-concept phase of GenAI use. The majority of CIOs are also seeking to generate new revenue (37%) or reinvent their business model (27%) by deploying GenAI.