Atrás 
This is a series of articles focused on Next Gen SWG use cases. This is the third in a series of six posts.
In my recent blog, I continued my blog series covering the most common use cases our customers are addressing with our Next Gen Secure Web Gateway (SWG). Next up is web filtering and coaching users on acceptable use.
If you are a seasoned pro when it comes to SWGs, your eyes may be rolling already. Web filtering and acceptable use have been the cornerstone of the SWG market since the early days of Websense and Blue Coat more than a decade ago. So why am I focusing on these seemingly well-understood and some would say, commodity capabilities? It is simple. The explosion of cloud app adoption, combined with the evolution of user expectations, is rendering legacy SWG features like web filtering and coaching ineffective in today’s digitally transforming world. Web filtering and user coaching needs to evolve to better meet today’s security requirements, which often have end user experience built in.
Here are some capabilities to look for in a Next Gen SWG when it comes to being able to deliver web filtering and user coaching modernized for today’s environment.
Machine learning-powered dynamic website classification
Being able to classify websites has been a table stakes capability of SWGs for more than a decade. Whether you are identifying malicious websites or looking for websites that have inappropriate content, understanding how to classify the website is the first step towards being protected and enforcing acceptable use policies. The technology used to classify websites has not changed much over the years and several vendors on the market specialize in website classification and maintain a database that they license to security vendors. While this method continues to be effective in covering a good percentage of web usage, the dynamic nature of the web is resulting in categorization misses. The result is you are at risk of threats and inappropriate content being delivered to your users. If you have configured policies based on website categories and the content does not match the category then bad things can happen.
A Next Gen SWG uses machine learning techniques to complement classification databases and improve the efficacy of website categorization. One example is looking at the body of the content of the URL being accessed and dynamically updating the classification if it is found to be incorrect, or dynamically classify websites that have yet to be covered by the classification vendors.
Comprehensive cloud app database with detailed cloud app risk rankings
When the SWG market first started more than a decade ago, cloud app traffic was a relatively small percentage of the overall web traffic. Today, more than 85% of today’s web traffic consists of cloud app usage. This presents a big gap for legacy SWG products as they provide a limited view of cloud apps, often combining the view with general web usage. Cloud apps are where the majority of threats persist and not getting a crystal clear understanding of cloud usage puts you at risk.
A Next Gen SWG should provide a robust cloud app database with more than 30,000 cloud apps and should provide ongoing research of each cloud app, assessing capabilities ranging from compliance certifications to data protection to understanding who owns data uploaded to the cloud app, the SaaS vendor or the user? Each cloud app should be assessed a risk score based on the ongoing research and the Next Gen SWG should support reflecting the risk score in policy enforcement.
SSL/TLS inspection at cloud-scale with native support for TLS 1.3
With more than 70% of internet traffic being SSL/TLS encrypted, the ability to inspect SSL/TLS traffic has been a common capability across SWG vendors. Relying on appliance-based SWGs to perform decryption can be problematic given the compute resources required. The result is customers end up purchasing more beefy appliances or, in some cases, dedicated appliances are deployed to do the job. Cloud-based SWGs take advantage of the cloud and can apply a virtually unlimited amount of resources to perform decryption.
There is a new TLS decryption challenge that has surfaced relatively recently, TLS 1.3. TLS 1.3 is a major revision to TLS 1.2 and one of the key capabilities is providing additional privacy for data exchanges by encrypting more of the negotiation handshake to protect it from eavesdroppers. This is problematic for man-in-the-middle security devices such as SWGs. The result is that many SWGs will force a negotiation from TLS 1.3 to 1.2 so they can inspect the traffic. This is not id