How CCTV, facial recognition, and AI may be forcing the discussion about data ownership

Publish date:

Who owns the data and how can the benefits be shared across the different actors of the data value chain.

About ten years ago, a friend of mine and I had a conversation about whether it was still possible for a person to disappear entirely. Go off the grid—no credit card, no traceable cell phone, no digital footprint whatsoever. That was at the very beginning of the smart phone hype and before debating cashless societies became a hot topic. Our conclusion back then was that no one could go off the grid entirely. There would always be some ICT-related string attached somewhere.

If my friend and I were to resume this conversation today, we would probably immediately reach the conclusion that disappearing is no longer possible at all. The brilliant demonstration made by the Chinese, who took just seven minutes to apprehend British BBC reporter John Sudworth, shows the power of mass surveillance[1].

[1] https://techcrunch.com/2017/12/13/china-cctv-bbc-reporter/

. And it’s not just about having set up millions of CCTV feeds. The data in itself is nothing. It’s the insight offered by the data that has value. There are three main takeaways from the Chinese hit:

  1. Whatever algorithm was used to filter and follow the reporter, this shows the mass application of AI for facial recognition. So, are you still okay about giving away your face for free to Apple when signing in on your new iPhone X? Huawei goes further, with the ability to capture 300,000 facial points in 10 seconds.
  2. Whereas CCTVs may still be far behind in terms of precision, it’s not so much the number of data points that matters the most, but rather the accuracy of a selection of data points combined with other data points, for example facial recognition combined with the color of the clothes, the licence plate number of a car, and so on. In this context, link and labelled data have a major part to play.
  3. Real-time data, real-time insight, and real-time action provide insight at the point of action. Live feeds were used and analyzed to follow and capture the reporter. Now, this is powerful stuff we’re only used to seeing in movies–proving the point that mass surveillance is reaching unprecedented new levels.

Dropping, for a minute, the conspiracy theory around mass surveillance, the applications for these types of features are massive, especially in the field of law enforcement and intelligence gathering. Quality controls could also be conducted remotely along the supply chain, identifying whether given hygiene or norms have been followed. Insurance companies may equally wish to invest in these types of surveillance solutions to protect property—be it houses, or parking garages in which alarms could be triggered in the presence of unfamiliar faces. Precision farming may also benefit from such surveillance solutions to further combine imagery with other sensors.

Is there a big difference between football coaches covering their mouth when giving out instructions to their team to prevent lip reading when playing back a recording of a given game and mass surveillance? You might say that the context, or the purpose, or the volume of data collected differ. So, how do we leverage AI in this context to develop resilient services?

I have two recommendations for 2018. These two recommendations revolve around ethics and symmetry. First of all, as ethics evolve they need to be redefined to match society’s evolution. This (re)definition is fundamental if we are to benefit from technological breakthroughs and sustainable societies. Secondly, symmetry is needed to balance the power between the data producer—aka the individual—and the data collector who makes use of the data. The fundamental question is: “who owns the data and how can the benefits be shared across the different actors the data value chain?”

Related Posts

AI

What to consider before implementing your RPA Center of Excellence (CoE)

Erlend Vist Ekren
Date icon October 19, 2018

A crucial step in moving from the proof-of-concept stage to distributing RPA across your...

agile

Enterprise’s valuable resource is data in the digital world

Shivakumar Balasubramaniyan
Date icon September 27, 2018

Data is a vital resource for an enterprise. Read this blog to learn more about the economics of...

AI

Don’t just replace human activities, make RPA transformational  

Ayan Ghosh Dastidar
Date icon September 25, 2018

Enterprises should focus on where and how an ‘RPA’ fits in the overall...

cookies.

By continuing to navigate on this website, you accept the use of cookies.

For more information and to change the setting of cookies on your computer, please read our Privacy Policy.

Close

Close cookie information