It's tough for businesses to make use of data as it grows in volume and complexity, pouring in from multiple sources and systems. As organizations use new technologies like artificial intelligence and blockchain to process and store data, they have a responsibility to ensure data is protected.
"How do we enable those workloads to transform the business, and to the best of our abilities prevent security from getting in the way?" says Rick Echevarria, vice president of Intel's Software and Services Group, and general manager of Platform Security, in an interview with Dark Reading.
At Cyber Week, now taking place in Israel, Echevarria will discuss how he's navigating the challenges in new tech. Machine learning algorithms often require access to sensitive data but limiting access to the necessary data could limit the potential for AI. As for technical considerations related to blockchain, the security of information is as important as scalability.
Getting Smarter on AI
When it comes to trends around AI, he says, there are two implementations: security for AI, where the focus is on protecting data and algorithms, and AI for security, in which tools leverage AI to detect advanced threats. Echevarria's focus is on security for AI.
"The reason for that is because people are deploying artificial intelligence today, and we want to make sure they're really aware of, and understand, the importance of security to underpin those workloads," he explains.
Echevarria points to two use cases in which Intel is addressing the AI challenge. The first is multi-party machine learning, which relies on the assumption that you can bring data to a central location. Tools like Intel's hardware-based Trusted Execution Environments grant access to critical data and ensure algorithms' integrity.
The second is federated learning, intended for applications where data can't be moved to a central location. In this case, data owners on the edges of the infrastructure work together to develop models themselves. Here, Echevarria says, many businesses encounter challenges.
"It's really about enabling access to data across organizations that really want to collaborate but sometimes because of privacy can't share data in the way they want to," he explains. The challenge is often seen in industries like healthcare and financial services, where teams may want to share threat intelligence but don't have the permission to share data.
Intel plans to support more extensive data sharing by teaming up with organizations like Docker, Fortanix, and Duality. With Docker, for example, a partnership will focus on making AI more secure and sharable by hardening containers with Intel's silicon-based security tech. Fortanix is adding to its Runtime Encryption tool to secure algorithms with Intel SGX enclaves.
Buckling Down on Blockchain
Some of the challenges in AI are consistent with challenges in blockchain, says Echevarria, who notes that Intel sees opportunity in blockchain as a platform. This week Intel is addressing scalability, security, and privacy of blockchain with updates in what it calls "Off-Chain Computing" along with partnerships with Enigma and SAP.
Intel has already teamed up with Microsoft to work on securing the blockchain. Back in May, the Microsoft Azure team, along with Microsoft Research, Intel, Windows, and the Microsoft Developer Tools division, announced it had been working to bring Intel's Trusted Execution Environments (Intel SGX, Virtualization Based Security) to the cloud.
Most recently, Intel partnered with SAP to build a blockchain proof-of-concept and improve efficiency for SAP's blockchain-as-a-service platform. Enigma has developed a privacy protocol that uses Intel SGX to secure data; it plans to integrate this into private smart contracts on the Ethereum public ledger.
Echevarria also speaks to plans to improve off-chain computing, in which transactions are not publicly accessible on the blockchain, specifically related to smart contracts. Businesses can do a better job of executing smart contracts in a more secure fashion, he says.
- Improving the Adoption of Security Automation
- The Best and Worst Tasks for Security Automation
- Most Websites and Web Apps No Match for Attack Barrage
- How to Prepare for 'WannaCry 2.0'
Why Cybercriminals Attack: A DARK READING VIRTUAL EVENT Wednesday, June 27. Industry experts will offer a range of information and insight on who the bad guys are – and why they might be targeting your enterprise. Go here for more information on this free event.