Amidst never-ending, highly publicised news reports of data breaches, hacker attacks, cryptocurrency theft, and other online scams against consumers and businesses, cybersecurity has once again moved to the top of the agenda and priority lists of many IT departments and providers worldwide.
It’s likely going to get worse before it gets better
We continue to witness high-profile Fortune 500 companies becoming victims of well-planned and executed breaches through cracks that left them vulnerable. The methods and toolboxes of cyber-criminals are becoming more sophisticated and well-funded. As we become increasingly connected through our many devices, the access points and surface for attacks also widen.
In an effort to stay ahead of not-so-well-intentioned online agents, the Cybersecurity Tech Accord was created and signed earlier this year. The accord is a public commitment among more than 30 global companies to protect and empower internet users and to improve the security, stability and resilience of cyberspace. It was signed by Microsoft, Facebook, SAP, and Cisco, among other tech giants.
"The devastating attacks from the past year demonstrate that cybersecurity is not just about what any single company can do but also about what we can all do together," said Microsoft President Brad Smith in an interview on the topic.
What are the top cybersecurity trends you need to be aware of this year if you are to keep things under control in your business? We’ll discuss the three we consider essential here.
The re-birth of video analytics and AI
Video surveillance is part of daily life in large cities and many offices. As AI grows more sophisticated and useful, capable of deciphering images, facial expressions and videos, the analytics produced by it are also growing in utility. Thousands of terabytes of video are recorded and stored on a daily basis around the globe as part of normal surveillance processes. Does anyone watch any of it? Hardly anyone ever does, unless there is a good reason for them to spend time looking for a specific event (e.g. a bank heist or another crime that needs to be investigated).
But what if we could derive intelligence from all these video feeds? What if we could tell which person, driving which automobile, ran which red light, or took part in which hit-and-run accident? These are all insights we can derive with the help of AI-driven video analytics that we are using already.
We could track and analyse much more, however. People tracking, area protection, anomaly recognition and reporting, suspicious or lost object detection are just some of the potential applications of advanced analysis. AI-driven video analytics are going to be a key technology for people, object and facility monitoring and emergency response in the coming years.
Solutions trump products
Small and mid-sized enterprises (SMEs) struggle to keep up with the latest securitytech, often citing lack of knowledge and resources as top concerns. To avoid having your vulnerabilities exploited by sinister organizations, it’s prudent to ensure your systems and response protocols are attack-proof and always up-to-date.
To accomplish this not-so-trivial feat, companies are looking for complete solutions, rather than products, to effectively protect their property, personnel, customers and assets, according to Roy Alves, sales director at Axis Communications. Rather than implementing products in silos, IT teams at large organizations and IT managers at SMEs, are in need of entire systems that can take care of security on all levels and help them tie loose ends.
Deep learning opportunities
Cybersecurity experts today are making use of both supervised and unsupervised deep learning to devise systems that can learn the nature of cyberthreats inside out, and thus, generate the appropriate responses, respective to each threat.
Deep learning is useful when detecting anomalies, like malicious behaviour or agents. The biggest challenge this field faces today is knowing what’s normal and acceptable and what isn’t. When it comes to normal and abnormal security events, there are exceptions to each rule. The tricky part is achieving a level of accuracy that would make the deep learning algorithm effective and practical enough to use in a real-world environment.
Although AI and deep learning are still far from solving all current security issues, it is also worth exploring applications that can make the work of security analysts easier and more efficient. Data visualization is one way of realizing this ambition and it is also useful when verifying and guiding the results of machine learning algorithms.
Copywriter: Ina Danova