Using Tech to Secure Tech

I was once wandering the streets of Vienna, Austria, with no local currency and no way to access my bank accounts. I had planned to get money out of an ATM once there, but my card wasn’t accepted. No problem: I had a credit card. Not having traveled overseas much, my credit card company decided I wasn’t me and blocked my card for my own good. I was hungry and jet-lagged. I was not impressed with my credit card company’s security algorithms.

But security technology, like any other technology, is constantly improving. In fact, it is likely to be improving faster than the public’s impressions of it. Research firm Gartner says the security posture of major cloud providers is as good as or better than most enterprise data centers. In fact, the firm predicts that through 2020, public cloud infrastructure as a service (IaaS) workloads will suffer at least 60% fewer security incidents than those in traditional data centers. But with data breaches making news with what seems to be increasing regularity, it’s understandable that many companies have second thoughts about making the leap into cloud computing or rolling out connected devices. No one wants to make news that way. DE’s audience is no exception. When asked what challenges they faced when developing internet of things (IoT) solutions, the most cited challenge (42%) was security. That’s down slightly from last year when it was cited by 48% of respondents—but it still beats out all design complexity and connectivity challenges.

Cultural issues beating out technical ones is a recurring theme we saw in our research. You can see survey results sprinkled throughout the issue. For example, survey respondents rated collaboration as their biggest day-to-day challenge at work, and many respondents voiced their distrust of artificial intelligence as an inhibitor to its adoption.

AI to Outsmart Hackers

Years after my overseas credit card mini crisis, my credit card was suspended again because it was being used by someone who was not me. The credit card company detected it, suspended it and sent me a new card along with tips on making sure my identity wasn’t stolen. The algorithms had improved, and they’re only getting better.

At Dell EMC World 2017 in May, Tony Parkinson, VP of North America Enterprise Solutions and Alliances at Dell EMC and Nick Curcuru, VP of the Big Data Practice at MasterCard International Inc. discussed how the companies are working together to stay one step ahead of cyberfraud with machine learning.

“For us, machine learning allows us to actually bring in the data so much faster,” said Curcuru, which allows MasterCard International to apply rules in milliseconds to understand which transactions are fraudulent and which are valid. “It’s not chasing; it’s ‘I am going to prevent this upfront if I can.’”

Curcuru says machine learning also reduces false positives so valid transactions by customers aren’t blocked. To do that successfully requires machine learning to “learn what you know so you can apply it to larger data sets,” he says.

From FinTech to Product Design

It’s not a huge leap to imagine how that type of machine learning could be used to make all sorts of online transactions more secure—such as simulation data and other intellectual property in the cloud or the data being collected by IoT-enabled devices. Many manufacturers have lots of data, but can’t gain any knowledge from it. Machine learning could help them learn what they know, which would then help them detect anomalies that might signal a security issue.

“A lot of verticals are interested in using machine learning and deep learning for security,” said Kash Shaikh, VP of Product Management & Marketing, Hybrid Cloud and Solutions for Dell EMC at last month’s supercomputing conference, SC17. “The key is accuracy. You want to be able to add security without generating false alerts.”

Assuming AI could allow companies to feel better about running simulations in the cloud, collaborating with others on cloud-hosted designs or rolling out connected products, there still leaves the cultural unease with AI itself. As one DE survey respondent wrote when asked about AI and machine learning: “Products will become more user friendly, easier to learn and more powerful, but this will take some time for humans to adapt and trust the AI systems.”