Digital Transformation With Cloud: Answering Risks With Algorithms

Digital transformation almost always demands cloud transformation. Even in 2019, however, cloud initiatives carry security risks that can't be ignored; cloud is simply too big for mere humans to tend to.

At least, without the help of machine learning.

"One of the biggest risks digital businesses face today is lack of visibility within their hybrid environments," Chris Smith, head of Global Security Services at CenturyLink, tells Security Now. "Today's IT departments are tasked to do more with fewer resources."

Sure, the whole point of digital transformation is to be able to do more with less, but you still need something in the way of OPEX and keeping the lights on. Many digital-transformation initiatives attract funding and other resources only for "sexy" parts, however -- with little regard for the behind-the-scenes stuff that makes the magic happen. Cloud transformation, therefore, necessitates cloud-security transformation. Otherwise, warns Chris Richter, an independent advisor for cybersecurity startups, CISOs and their brethren will lag.

"If [a company is] moving to the cloud, they need to look at the applications that are moving to the cloud and ensure that what's going there is secure," says Richter. "While the innovators in organizations are pushing for digital-transformation, data-driven services [like cloud], security to go along with it is falling behind in keeping up."

It's my third party so I'll cry
The same goes for third parties. Cloud has so penetrated the enterprise market that, as the US Department of Treasury has pointed out, we have been seeing the entirety of an enterprise's critical functions get outsourced -- at times without proper risk assessment, auditing, due diligence, or even contractual safeguards. Sometimes, too, the first-party organization may get the wool pulled over its eyes by cloud vendors who over-promise and under-deliver -- particularly if those vendors also offer encryption services that dictate that they control the cryptography keysto the client's data instead of the client him/her/itself.

And all of these problems are exponentiated when fourth parties (third parties' subcontractors) get introduced, making for yet more potential attack vectors and data exposures, yet more due diligence to be conducted, and yet more jurisdictional and regulatory headaches based on the digital transformation extending to data geography.

This is particularly problematic when the third-party cloud transformation extends into functions that are customer-facing and/or involve customer data, exposing the first-party organization to yet greater and more costly risk. Repeatedly, third-party data exposures and breaches have been a top cloud-transformation concern among IT pros -- and, repeatedly, third-party (and first-party) data exposures and breaches happen in cloud environments. The past two years in particular have seen numerous headline-splashing incidents of AWS cloud S3 buckets left unsecured -- a non-default setting, meaning that an administrator almost certainly had to purposely go out of their way to screw things up.

People problems
Effectively, this all comes down to human error. Security experts point out that human decisionmakers are too quick to assume such things as that their contractors are trustworthy, that their on-prem security measures automatically translate to effective cloud security, or that nobody will take advantage of their opsec flubs. There are even instances where mathematically challenged teams have so mismanaged their cloud resources as to effectively create their own DIY ransomware-as-a-service.

"It's not for lack of trying [or] laziness; it's just the fact that they are moving so fast that they overwhelmed with the vulnerabilities that emerge," said Richter as he highlighted the benefits of machine learning to help ease the security pains of digital transformation. "We need to get to the point where humans are removed from the configuration question. Humans can be used to validate and authenticate an infrastructure, but we need to get much better at using machines [and] artifical intelligence to lock down [and check] an environment -- because most of these vulnerabilities if not all of them... can be traced back to mistakes that a human made." (See: Unknown Document 749545.)

Of course, employees resist digital-transformation efforts precisely because of their potential to leverage automation to eliminate jobs. Gartner, however, recommends cybersecurity and data-privacy AI tools to aid in digital transformation efforts as "augmented intelligence" -- aiding human efforts without wholesale replacing the human. (See: Privacy & AI Changing the Digital Transformation Game and Cybersecurity AI: Addressing the 'Artificial' Talent Shortage .)

Here, though, the issue rather becomes one of trusting a third-party product instead of a third-party service. Ultimately, humans must be accountable.

Richter acknowledges the point, but points out that the human decisions of managing cybersecurity risk in digital transformation should remain at a high level only -- leaving the nitty-gritty day-to-day operations to the machines. And, indeed, some companies are offering specialty automation and machine-learning tools specifically as cloud-security solutions. Meanwhile, tech pundits like writer Wayne Rash argue that, despite the risks, cloud vendors are probably better able to handle your data and infrastructure than you are. (See: Sophos & Akamai Target Cloud Security With Acquisitions.)

After all, Jeff Bezos didn't unsecure all those S3 buckets.

Related posts:

— Joe Stanganelli is managing director at research and consulting firm Blackwood King LC. In addition to being an attorney and consultant, he has spent several years analyzing and writing about business and technology trends. Follow him on Twitter at @JoeStanganelli.