Threat Intelligence

9/19/2017
02:30 PM
Terry Ray
Terry Ray
Commentary
Connect Directly
Twitter
LinkedIn
RSS
E-Mail vvv
50%
50%

GDPR & the Rise of the Automated Data Protection Officer

Can artificial intelligence and machine learning solve the skills shortage as the EU's General Data Protection Regulation deadline approaches?

For many companies, May 28, 2018, will be a watershed moment for data security as the General Data Protection Regulation (GDPR) comes into effect with the intent of strengthening and unifying data protection for all individuals within the European Union (EU). Companies throughout the EU (and the world) must be prepared to comply with the regulations or suffer severe penalties for the inappropriate protection or handling of customer data.

To oversee the data, the GDPR calls for companies processing personal data on a large scale to retain an independent data protection officer (DPO). The DPO in essence becomes the "voice" of data protection and compliance within the company.

The DPO requirement spurs one of the most pressing questions related to GDPR: Who is qualified to fill this role when there aren't enough professionals with a security background to fill existing positions within the industry? A study conducted in April 2016 by the International Association of Privacy Professionals (IAPP) suggests that at least 28,000 DPOs will be needed across Europe to meet compliance requirements. When thinking globally about organizations that operate with EU-resident data sets, that number is likely to rise to somewhere in the region of 75,000, meaning a serious lack of qualified data protection professionals is likely, if not unavoidable.

Some within the technology industry have begun to suggest that artificial intelligence and machine learning can play a role in assisting DPOs. However, using machines that are capable of learning and carrying out the complex processes necessary for data protection could be controversial when considering the sensitive data involved. Before assessing the likelihood of this, it is important to understand the DPO role and job function.

Under GDPR, the DPO is an independent advocate for the proper care and use of EU citizenry data. The job requires them to remain current regarding data protection laws and practices, conduct internal data protection impact assessments, and ensure that all other data compliance matters are up-to-date. Typical DPO projects will include but won't be limited to: data retention, data anonymization, security risk assessment of business practices involving personal data, privacy impact assessment of new products and services, platforms, vendor assessments and audits, Internet of Things, and breach management. 

This will be a labor-intensive role. Yet in a recent survey of 310 IT security professionals conducted by Imperva at the 2017 Infosecurity Europe conference, more than half (55%) of respondents indicated that they believed AI or machine learning solutions could bear some of the DPO's considerable workload.

Because most organizations that will be affected by GDPR will be working with large data sets, these sets lend themselves naturally to a machine learning- or artificial intelligence-based solution. Machine learning technologies are incredibly adept at analyzing large data sets and establishing patterns of both good and bad behavior. Therefore, as opposed to a data protection officer employing several experts to analyze trends and patterns within the data, a machine learning solution could be deployed to do this work at a much faster pace, more thoroughly and ultimately more cost effectively.

These technologies can also help to establish an automated response to one of the most prevalent threats to an organization's data: insider threats. While the motivation behind insider threats can be varied — some users are simply careless or negligent, while some have genuine malicious intentions to sell the data for profit — all insider threats are likely to seem anomalous when accessing data inappropriately.

Specific artificial intelligence techniques, such as natural language processing, could provide solutions to some of the problems of sensitive data that needs disposal. Natural language processing has in the past been used in other industries, such as law, to scan vast sums of documentation to see which ones appear relevant to a case. This same technique could be used to establish whether data has expired, or if it is no longer relevant.

In addition, artificial intelligence and machine learning solutions have the power to affect the role of the DPO significantly, incorporating the ability to assess and analyze vast sums of data in a fraction of the time it would take for even the most talented security professional.

However, as is so often the case when discussing the benefits of human labor vs. AI, there are some tasks where the levels of inference are simply too complex to be completed by an automated DPO system … for now, at least. This could include related management tasks, such as data breach protocol, providing recommendations to meet compliance standards, and feeding back data-related queries to other members of the organization.

The most pragmatic solution would be to hire a DPO who is comfortable working with data and security, and who also has a background in the legal structures surrounding data compliance. But despite the great promise of AI, businesses that handle European data cannot yet place their faith in machines to ease the burden of the DPO and meet the pending requirements of the GDPR.

Related Content:

Join Dark Reading LIVE for two days of practical cyber defense discussions. Learn from the industry’s most knowledgeable IT security experts. Check out the INsecurity agenda here.

Terry Ray has global responsibility for Imperva's technology strategy. He was the first US-based Imperva employee, and has been with the company for 14 years. He works with organizations around the world to help them discover and protect sensitive data, minimize risk for ... View Full Bio
Comment  | 
Print  | 
More Insights
Comments
Newest First  |  Oldest First  |  Threaded View
martin.george
50%
50%
martin.george,
User Rank: Apprentice
9/25/2017 | 11:08:23 AM
Great post
Well, it is really pretty great post - I haven't seen such for the long time 
WebAuthn, FIDO2 Infuse Browsers, Platforms with Strong Authentication
John Fontana, Standards & Identity Analyst, Yubico,  9/19/2018
Turn the NIST Cybersecurity Framework into Reality: 5 Steps
Mukul Kumar & Anupam Sahai, CISO & VP of Cyber Practice and VP Product Management, Cavirin Systems,  9/20/2018
NSS Labs Files Antitrust Suit Against Symantec, CrowdStrike, ESET, AMTSO
Kelly Jackson Higgins, Executive Editor at Dark Reading,  9/19/2018
Register for Dark Reading Newsletters
White Papers
Video
Cartoon Contest
Write a Caption, Win a Starbucks Card! Click Here
Latest Comment: "I'm not sure I like this top down management approach!"
Current Issue
Flash Poll
The Risk Management Struggle
The Risk Management Struggle
The majority of organizations are struggling to implement a risk-based approach to security even though risk reduction has become the primary metric for measuring the effectiveness of enterprise security strategies. Read the report and get more details today!
Twitter Feed
Dark Reading - Bug Report
Bug Report
Enterprise Vulnerabilities
From DHS/US-CERT's National Vulnerability Database
CVE-2018-17332
PUBLISHED: 2018-09-22
An issue was discovered in libsvg2 through 2012-10-19. The svgGetNextPathField function in svg_string.c returns its input pointer in certain circumstances, which might result in a memory leak caused by wasteful malloc calls.
CVE-2018-17333
PUBLISHED: 2018-09-22
An issue was discovered in libsvg2 through 2012-10-19. A stack-based buffer overflow in svgStringToLength in svg_types.c allows remote attackers to cause a denial of service (application crash) or possibly have unspecified other impact because sscanf is misused.
CVE-2018-17334
PUBLISHED: 2018-09-22
An issue was discovered in libsvg2 through 2012-10-19. A stack-based buffer overflow in the svgGetNextPathField function in svg_string.c allows remote attackers to cause a denial of service (application crash) or possibly have unspecified other impact because a strncpy copy limit is miscalculated.
CVE-2018-17336
PUBLISHED: 2018-09-22
UDisks 2.8.0 has a format string vulnerability in udisks_log in udiskslogging.c, allowing attackers to obtain sensitive information (stack contents), cause a denial of service (memory corruption), or possibly have unspecified other impact via a malformed filesystem label, as demonstrated by %d or %n...
CVE-2018-17321
PUBLISHED: 2018-09-22
An issue was discovered in SeaCMS 6.64. XSS exists in admin_datarelate.php via the time or maxHit parameter in a dorandomset action.