Intel's confidential computing is poised to tackle growing enterprise security concerns
With a shifting regulatory environment and a constantly evolving threat landscape, security is never far from boardroom discussions.
The challenges for many are simple: how can they glean more value from data and new technologies without compromising security standards? In highly regulated industries, this balancing act translates to siloed off datasets that are intrinsically valuable but cannot be processed using new capabilities. Perhaps the best example is confidential patient data which, when processed at scale using cloud technology, could help track the evolution and spread of diseases - but cannot because of stringent privacy regulations and security concerns.
Through confidential Computing, however, this challenge is quickly remedied. According to Intel, who are pioneering solutions in the space, confidential Computing enables encrypted data to be processed in memory while limiting access to the rest of the system, thereby reducing the potential for sensitive data to be exposed while providing a higher degree of control and transparency for users. Intel's Software Guard Extensions (Intel® SGX) capability enables this by encrypting memory to form enclaves; applications can safely access and decrypt the data inside an enclave if they have the necessary access keys. This means whether the data is processed on-premise or in the cloud, users can exert more protection controls over their IP and data.
Simply put, this technology enables businesses to secure their most prized IP and data within the hardware itself during processing. By doing so, they can share overlaying products and services – with customers and partners, for example – without compromising the security of the data underneath. Because the encryption is embedded within the hardware, it presents the smallest attack surface possible, with the capacity to rapidly freeze out and quarantine any anomalies or threats before they impact the broader technology estate.
Intel boasts a plethora of use cases and examples of successful implementations. Sticking with the healthcare sector, the firm worked with The University of California San Francisco's (UCSF) Center for Digital Health Innovation to ensure confidential data is protected when sharing developed AI solutions with partners. Part of the process of developing breakthrough medical devices and capabilities is to have the data and algorithms validated by partners. With confidential patient data locked away at the microprocessor level, UCSF is able to work more closely with partners to certify solutions without the risk of exposing valuable data or IP.
Similarly, Intel partnered with Magnit, a leading Russian supermarket chain in the retail sector to protect confidential consumer data. A measure necessary when working with partners on engagements such as advertising programmes. Again, by securing vital data at the processor level, the firm can protect its IP and assure privacy and security without compromising on the need to collaborate with external partners.
While use cases are clear - with cross-industry relevance, but a particularly strong proposition for highly regulated industries - there are still challenges articulating the value of confidential computing approaches. According to Richard Curran, Intel's Datacenter Security Officer, IT Executives may not recognise the term Confidential Computing, but they are certainly aware of the challenges Data Privacy imposes and are excited by this initiative.
Take, for example, the desire to benefit from the economies of scale from cloud computing. Companies that cannot trust their prized IP and Data outside of their own data centres simply can't exploit the technology. But with greater assurance provided by Confidential Computing, they can. While this sounds like a simple adjustment, the results can be considerable. For some large institutions, right now, real security assurance means they have to spend millions on infrastructure through a capital-intensive business model that enables them to analyse data internally in their own data centres. Confidential Computing enables them to take advantage of Cloud economics, thus driving efficient, agile IT business models and accelerating business transformation.
According to Curran, this simple shift enables greater collaboration across industry - simply because enterprises can lock down the critical data and IP, without inhibiting their ability to share the overlaying product. A move that will likely grow in importance as we see more connected devices, and subsequently more data, spread across the global economy. At one end, it powers the secure sharing of data across financial institutions to support anti-money laundering efforts. And at the other, it enables the harvesting and processing of data from self-driving cars without the immediate need to tackle initiatives such as deidentification, which hampers the speed and quality of development.
As the world becomes more connected, protecting IP is as vital as safeguarding valuable data. According to Paul O'Neill, a Strategy Lead in Intel's Confidential Computing Group, as enterprises push more compute and processing capabilities to the edge of their enterprise, they are also effectively shipping valuable IP - such as algorithms to build federated models for machine learning techniques. Ensuring this IP is secured is critical. It's this development, according to O'Neill, that will see more emphasis on confidential computing technology from product and data owners looking to secure valuable IP to drive insights from data, alongside traditional stakeholders such as CIOs and CISOs.
With more enterprises under pressure from customers, regulators, and competitors to provide greater security, it's likely we'll see confidential computing move further up the enterprise agenda. For some, incorporating the technology into their products and services will form part of a vital differentiator. For others, it will be a necessary evolution of their secure-by-design and zero-trust approaches. In any case, we can expect the ongoing development in the space from firm's such as Intel to gain more traction as enterprises push for innovation in the security space.