It’s a race - on one side: fast moving technologies, new business practices, new digital behaviours, the democratisation of analytics and the massive adoption of artificial intelligence (AI) and machine learning technologies. On the other side: the fundamental right of individuals to protect their privacy. On one side: what we can do, what is technically possible. On the other side: what we should do, what is ethical.

It’s a race between data-driven innovation and privacy, and it’s fair to say that there was some catching up to do. When it comes to privacy and personal data protection, the previous directive was drawn up in 1995. That was before the massive adoption of the internet, social networking, and e-commerce. Since then, the world has changed. It has become data-driven, and personal data is the new gold.

• Private organisations, large and small, started gathering as much consumer data as they could to fuel new data uses, create insight and make real-time, fact-based, automated decisions that have a direct impact on our lives: to determine whether one should get a loan, whether one is likely to be criminal, whether one is likely to buy something and at what price.
• Political parties are using data and analytics to win elections.
• Governments are using data and analytics to tackle tax fraud, or to identify children who may be at risk of abuse.
• Hospitals are using machine learning techniques to assist with cancer diagnosis.
• Football clubs are using video analysis to inform decisions on tactics, lineups, training schedules, evaluating the true potential of players and predicting match outcomes (see Hidden Insights article by Reece Clifford: “How Analytics Is Shaping the World of Sport”).
• Individuals have become data-driven, too, using vast amounts of data available on the internet to make buying decisions, using AI every day to book holidays, to look for a job, to find a life partner, in their phones, in their cars, etc.

Personal data is everywhere and highly accessible, and technologies are making it easy to use this data. As a result, privacy has never been as exposed!

Awareness is building up too. Massive data breaches and revelations about the inappropriate use of personal data, or the lack of protection given to personal data, are bringing the question of privacy to the headlines.

With the now enforced EU GDPR (Global Data Protection Regulation), the planets are aligning for a complete revamp of what privacy means, how personal data is handled, and what it means for organisations and their ability to innovate with data.

During Analytics Experience on 22 – 24 October in Milan, Olivier Penel will dive into the challenges and opportunities that fuel innovation in a compliance-focused world.

Nowhere to hide!

Innovating with data brings a lot of benefits to companies, people and society in general. But it can also expose the privacy of individuals in unprecedented ways.

An illustration of this is given in the documentary Pre-Crime, which shows how police enforcement authorities in the US are using large amounts of data and machine learning algorithms to predict the likelihood of somebody committing a crime or to identify places and times where a crime is likely to be committed. Surely, preventing a crime from happening is better than having to deal with the consequences of the crime.

But what if a person is wrongly targeted with preventive measures? Tagging someone as a potential high-risk criminal can have dramatic impacts on his or her life. Who is accountable for making that decision? What logic did the program use to reach that conclusion? What data was used, and is this data accurate? How do we correct the data and get the person off the list?

The question of accountability is even more critical when using machine learning algorithms. The developer who coded the program doesn’t know anything about crime and socioeconomic and psychological factors that lead to crime. He doesn’t even know what the program will eventually do as the program is designed to change over time, “learning” from the data.

Another example of disruptive innovations that raises privacy concerns is the recent breakthrough in face-recognition technologies. Face recognition is not new. It has been used for some time now, in some places such as border control, or by Facebook to flag people in our photo albums. But face recognition is now becoming mainstream, with exponential adoption across many domains and industries.

In parallel, the technology itself is getting better and better every day. For instance, research has shown that it is possible to reconstruct the face of a person based on his or her DNA, which can be very useful when it comes to identify the victims of an accident, or to catch the criminals using the DNA left behind. The same techniques can be (ab)used by authoritarian regimes to monitor citizens whereabouts and track down political opponents.

In the same way as driving a car safely requires drivers to learn and respect the code of conduct and rules defined by society for road usage, innovating with data requires organisations to learn and respect the rules and regulations designed to protect the privacy of individuals.

Evening out the game

Personal data is and remains a valuable asset. Organisations that manage to find the balance and stop the tug of war between innovation and data privacy will build a significant competitive edge.

Privacy-friendly analytics requires the ability to manage and control the use of personal data across the end-to-end analytical life cycle: from data discovery and visualisation to data preparation, model development, deployment and automated decision making. Privacy must be included at every step of the cycle, by design and by default. To achieve this, organisations who are looking to use personal data as a source of insight and competitive differentiator will need a complete and mature analytics platform.

If you want to learn more about this topic, join the conference Analytics Experience – Analytics redefines innovation on 22-24 October in Milan. Please register here for your seat.