The Internet Has Changed

Over the last decade, the internet has undergone significant changes. While in the past the internet was primarily used for accessing information, in recent years it has become a connectivity tool, providing access to critical business SaaS and IaaS. These cloud-based services have enabled businesses to be more agile and flexible, supporting remote work and vastly improving collaboration with colleagues across the globe. 

But while the internet—and our use for it—has changed dramatically, the security tools we place between our users and the internet (historically the web proxy) haven’t changed anywhere near as much. Here are five ways in which the internet has changed, with thoughts on how security needs to adjust to fit.

1. Critical business services are on the internet

You’d struggle to find any business today that isn’t using some form of cloud app. They are now so critical to the running of any business that I prefer to avoid the cutesy “cloud app” name and call them critical business services. But despite their ubiquitous presence, they exhibit huge diversity in the compliance and data protection standards in place.

We track the privacy and security characteristics of more than 60,000 different cloud apps that regularly pop up in use among businesses, and at the ‘most trusted’ end of the scale you find a few popular SaaS services that almost all businesses use to manage their day-to-day operations. I am talking about the likes of Microsoft 365, Amazon Web Services, Google Suite, and Salesforce.

They carry big, trusted names, and as such are often regarded as safe—sometimes to the point where they are allowed to entirely circumvent internet security.  

There’s two points to take away here. Firstly in this changed internet, we need security systems that have the ability to provide accurate visibility, control, and reporting of all cloud applications, including insights into the varying risks they carry. Old school web proxies just cannot do this.   

Secondly, we need to rethink the allowances we make for trusted brands. Microsoft cannot assure that OneDrive links do not host nefarious content in the way that they can make guarantees about their website. While in the past we may have dedicated ports and allowed lenient policy exceptions for certain outgoing application traffic or URLs, this is a dangerous approach now that the internet has changed. All internet-based applications – regardless of the brand they bear – must be included in stringent security policy. And legacy security often struggles to do this without severely hampering user experience.

2. Our infrastructure is on the internet

It isn’t just SaaS… organizations are also migrating their infrastructure into IaaS cloud providers like Amazon Web Services (AWS), Microsoft Azure, and Google Cloud Platform (GCP). 

When this infrastructure is no longer inside a perimeter we need to ensure that visibility and controls can be maintained and, again, traditional web security struggles. It’s incredibly common for organizations to consider IaaS another exception—essentially circumnavigating security in the interests of data speeds—for the crucial central infrastructure which arguably needs securing the most. Security needs to provide the same visibility and control of all the various resources (scripts, containers, and kubernetes) within IaaS, as well as the actions taken by their administrators inside these services and resources. 

3. Identities and instances are confusing on the internet

Not all OneDrive access is created equally… but security and data protection technologies built for a world before the internet changed struggle to understand this. 

It is normal for an organization’s employees to have a legitimate need to access multiple instances of the same application – perhaps a partner’s or customer’s OneDrive, perhaps a personal OneDrive. Instance detection is critical here, helping businesses identify which instance of an application an employee is interacting with, and designing nuanced and granular policies based not just on the URL but also on account and instance too.

To illustrate, without instance detection, when a folder is shared from one organization’s corporate Google Drive to another, there is a risk that sensitive information from one organization can accidentally be shared with the other just because outdated security systems think Google Drive is always fine.

4. Permissions and “segregation of duty” on the internet

The internet used to be a place where you consumed information—and security was built with that use case in mind. But cloud applications are not like traditional websites. They have a vast array of sophisticated capabilities that users can access and allow; ranging from editing and creating, to deleting and sharing files. As such, businesses need to be aware of the multiple different activity commands that users can trigger, and be able to provide accurate permissions to ensure proper “segregation of duty.”

5. Bad URLs look like good URLs

Do you remember when a large part of an annual security training course included tips such as “Look at the URL for clues that something dangerous lurks there”? This is yet another way in which the internet changed, in that it is now often much harder for a user to apply a basic set of static rules to avoid risky activity. The fact is that nearly half of all malware downloads now come from the cloud, and so it’s more important than ever that security tools assist users in avoiding risk.

Nuanced policy needs to work hand-in-hand with real-time coaching to ensure that users can continue to participate in defense. You can read more about real-time coaching in this recent blog post, but in essence it provides users with real-time feedback, and guidance on how to safely use cloud applications. Rather than simply blocking access to certain websites or resources, users can be educated on the potential risks and how to mitigate them. 

While security controls such as “blockpages” do still have their place in securing cloud applications, real-time coaching and security awareness provide a much more effective and proactive approach. By providing users with the necessary knowledge and tools to protect themselves, and make decisions with a clear understanding of risk, organizations can reduce the likelihood of security incidents and create better digital citizens in the process.

The internet has changed and so should you

As the internet has evolved, we too need to adapt. The internet today is full of nuance and context, and it requires security technologies that speak the language of cloud (JSON) and allow for granular policy application. Web proxies and legacy secure web gateways are unable to strike the balance between enablement and control, but a modern, intelligent security service edge stack can ensure organizations can mitigate risk (in an evolving threat landscape), while also running fast to embrace the benefits of cloud.
Read more about Security Service Edge in our Dummies Guide.