The Up and Coming
We’ll be starting a new blog segment evaluating up and coming market trends in VC. When evaluating potential market trends, we looked to those close to or currently being monetized, as well as new and exciting sub-sectors of old trends. For this week, we’ll be looking at Artificial Intelligence Applied to Cybersecurity.
An Emerging Market?
First, let’s look at the potential in cybersecurity. Cyber attackers can currently leverage automation to launch strikes in seconds, while organizations are responding manually, which takes weeks to months. Traditional countermeasures such as firewalls and antiviruses are no longer equipped to slow down hackers, much less stop them. In 2015, a study by PwC found that 79% of respondents detected security incidents during the year. Another study by the University of Singapore reported that security breaches cost victims more than $500 billion each year worldwide. These vulnerabilities are a result of a global shortage of experienced security workers and the expanding data sets from the rise of IoT devices. There are “wider and deeper” attack surfaces to protect, (ie. Cloud services, mobile devices, smart devices, applications) an amount projected to increase from 15 billion to 50 billion by 2020.
So where does AI come in?
Artificial Intelligence has become the buzzword of the startup world, often touted as the key to unlocking the future. So how does AI play into the Cybersecurity industry? Not surprisingly, AI integrates perfectly to address current Cybersecurity issues. Artificial intelligence and machine learning can be used to improve detection and response capabilities, handling large data sets quickly and on a 24/7 basis. The introduction of AI into cybersecurity reduces the standard workload of security analysts, so that humans can focus on new categories of threats. By better identifying suspicious patterns and behavior, AI integration can build a framework that is more resilient and adaptable on its own. With artificial intelligence, one data scientist can potentially replace more than 100 security analysts.
How It Works
Generally, artificial intelligence cybersecurity systems work like the human immune system: identifying threats, neutralizing them, and self-learning throughout the whole process to take on new vulnerabilities. There are three general steps:
1. Identification of Threats: The “wider and deeper” attack surfaces mean more feeds from disconnected systems that must be analyzed. AI systems aggregate the volume of security data to pinpoint needles in the haystack
2. Risk Assessment: AI determines the risk of security gaps and impact on the business, then identifies appropriate responses to individual risks.
3. Remediation: Finally, the AI system collaborates with human teams to remediate issues
Current solutions have focused on automatically sifting through data and patterns to identify threats and facilitate a human-led remediation plan. One risk associated with AI applied to Cybersecurity is Alert Fatigue or unsupervised machine learning leading to too many false positives and alerts. However, future solutions will seek to mitigate this risk and further entrust AI with preemptive threat mitigation and auto-remediation.
Market Trends and Exits
CB Insights analyzed millions of media articles to track mentions of cyber security, artificial intelligence, and machine learning, showcasing a steep incline over the last year. In 2016, there was a record of over 400 deals in the Cybersecurity market and 8 Cybersecurity Unicorns valued at over $1 billion including the AI Cybersecurity star, Cylance.
This table provides a brief overview of the fastest-growing AI Cybersecurity startups with already massive valuations in the span of a few years. Cylance, one of the first AI Cybersecurity startups to appear on the scene, was only founded in 2012 and is growing at over 1,000%.
In the acquisitions market, big players such as HP and Amazon have also been looking to beef up their cybersecurity by acquiring smaller AI Cybersecurity startups.