Orange Silicon Valley hosts a range of programs through which our analysts research and engage with emerging technology trends. Cybersecurity is one of those programs at OSV, and the research in this area focuses on one simple objective: protecting systems from cyberattacks.
Cybersecurity today is more relevant than ever before with security incidents rising in magnitude and frequency. In this environment, regulators have acted to protect consumer privacy by introducing new legislation, such as the European Union’s General Data Protection Regulation (GDPR) and California’s Consumer Privacy Act. However, as technology progresses, the threat landscape constantly evolves and grows.
Large financial institutions receive around 200,000 cybersecurity-related alerts each day, according to survey data from the research firm Ovum. That high volume makes it increasingly difficult for human analysts to identify actual threats. In this context, companies are increasingly leveraging machine learning to better distinguish between good and bad actors and identify malicious behavior or entities, in addition to automating menial tasks. However, identifying malicious activities is difficult because defining normal cyber behavior proves to be a big challenge. In order to better contain attacks, the cybersecurity community has turned to proactive approaches. Specifically, practitioners are designing better security architectures and principles so that the systems are more secure by construction rather than relying on machine learning algorithms to defend them later on.
“The cybersecurity market is fragmented with most entrepreneurs solving specific point problems and building feature companies around their solutions. Although Chief Security Officers want products that mitigate risk, they also want to reduce the number of vendors they rely on and improve the orchestration of the products they use. Therefore, we see the space for orchestration technologies as one of the many promising investment opportunities.”
Alex Chitea, Cybersecurity Principal
Broadly speaking, there are two machine learning algorithms used in cybersecurity: supervised and unsupervised techniques. Supervised algorithms is the area where machine learning has made the biggest impact (e.g., malware detection). Using this technique, an algorithm is trained on the different kinds of threats a system has faced before, and it is expected to catch them in the future. However, cybersecurity data is highly unbalanced by nature, and a very small number of attacks are buried in a sea of normal activity — which is where unsupervised learning is useful. Unsupervised learning techniques look at massive amounts of unlabeled data and find the pieces that do not follow typical patterns. As a result, these algorithms can surface threats that a system has never seen before, and they need few anomalous data points to do so.
As cybercrime penetrates more aspects of our daily lives, an increasing number of areas are rising in importance.
One sector is hybrid cloud security. Corporations are increasingly hosting their production applications in the cloud and operating across hybrid cloud infrastructures. This means that sensitive data is often housed in both public and private enterprise cloud environments, which increases the security perimeter and associated risks.
Another area is automotive cybersecurity. Automobiles are becoming connected and autonomous which makes them vulnerable to cyber-attacks. As a side effect, we expect insurance models to shift from covering the decreasing number of car accidents and instead focus on data breaches and accidents that result from security incidents.
Finally, there is the healthcare industry, which together with financial institutions is the most attacked vertical. Hospitals and internet-connected medical equipment are increasingly vulnerable to cyber-attacks, especially ransomware.
At Orange Silicon Valley, our cybersecurity program assesses ways to strengthen the cyber landscape to better protect users while enabling business services for other corporations. To learn more about cybersecurity research at OSV you can visit our Cybersecurity program homepage. Additionally, there will be a half-day cybersecurity conference led by Orange Silicon Valley Cybersecurity Principal Alex Chitea on the future of data security at the Silicon Valley Executive Summit on June 27th.