566% increase of compromised records in 2016
For many years, organizations have focused their security efforts on endpoint protection. Firewalls, antivirus software, intrusion detection and anti-spyware tools are all effective to a point, but they are failing to stop the vast majority of threats.
A recent ServiceNow survey of 300 chief information security officers found that 81% are highly concerned that breaches are going unaddressed and 78% are worried about their ability to detect breaches in the first place. IBM’s 2017 X-Force Threat Intelligence Index reported a 566% increase in the number of compromised records in 2016 compared to the previous year. FireEye reported that the average time it takes an organization to detect an intrusion is over 200 days.
Endpoint security measures will only become less effective as the number of endpoints proliferates. Smart phones introduced a whole new class of threats, and the internet of things (IoT) will add billions of endpoint devices to networks over the next few years, many of which have weak or no security.
That’s why cybersecurity, in the words of Cisco CEO Chuck Robbins, “needs to start in the network.” The approach that Cisco is championing recognizes the reality that breaches today are inevitable but that they needn’t be debilitating. The increasing popularity of security operations centers shows that IT organizations are shifting their attention to creating an integrated view of all the activity on their networks – including applications, databases, servers and endpoints – and adopting tools that can identify patterns that indicate a breach. For example, multiple access attempts from a certain IP address or large outbound file transfers may indicate an intrusion, and that activity can be stopped before much damage is done.
Fortunately, technology is evolving to support the network-centric approach. Big data platforms like Hadoop have made it practical and affordable for organizations to store large amounts of data for analysis. Streaming platforms like Apache Spark and Kafka can capture and analyze data in near real-time. Machine learning programs, when applied to large data stores like Hadoop, can continuously sort through network and server logs to find anomalies, becoming “smarter” as they go.
And the cloud presents new deployment options. That’s why security is rapidly migrating from dedicated hardware to cloud-based solutions using a software-as-a-service model. Grandview Research estimates that the managed security services market was worth more than $17.5 billion in 2015, and that it will grow to more than $40 billion in 2021. As organizations increasingly virtualize their networks, these services will become integrated into basic network services. That means no more firmware upgrades, no more site visits to fix balky firewalls and no more anti-malware signature updates.
It’s too early to say that the tide has turned favorably in the fight with cyber-criminals, but the signs are at least promising. It’s heartening to see Cisco making security such important centerpiece of its strategy. Two recent acquisitions – Jasper and Lancope – give the company a prominent presence in cloud-based IoT security and deep learning capabilities for network and threat analysis. The company has said that security will be integrated into every new product it produces going forward. Perhaps that’s why Robbins has called his company, “the only $2 billion security business that is growing at double digits.”