Does GDPR protect the privacy of your emotions?
Amazon Web Services announced this week that its facial recognition software, Rekognition, can read your emotions. Is that allowed?
Facial recognition is one of the most controversial artificial intelligence technologies around these days with debates already raging across the world.
It should come as no surprise that China’s authoritarian government has driven the most extensive and widespread facial recognition usage as part of its national biometric identity programme, storing a large proportion of its 1.4 billion citizens’ facial scans. Meanwhile, Chinese schools and universities are embracing facial recognition gates for security and facial recognition is revolutionising retail payments, with China’s tech giants investing heavily in developing face scan payment solutions.
In Europe, although there’s an arguably higher level of concern for the privacy, personal freedom and human rights considerations, technology firms and organisations are piloting facial recognition for various purposes. However, in the West, this technology is being implemented with caution.
This year, British police have pushed back against Home Office moves to try to roll-out facial recognition systems across all police forces. Two county police forces were conducting trials of the technology to find missing persons, but the news fueled privacy fears and created such a public backlash that British police, collectively, slammed on the brakes.
Meanwhile, in Sweden this week, the Swedish Data Protection Authority has issued its first ever fine under the EU’s General Data Protection Regulation, more commonly known as GDPR. The Authority fined the municipality of Sweden’s north-eastern Skelleftea city 200,000 Swedish Krona ($20,700) for conducting a three week facial recognition trial on high-school students in order to keep track of attendance. In doing so, it was found to have contravened privacy laws.
Facial recognition has a broad variety of applications for government and commercial use including providing a secure ID for payments, improving access and security, identifying criminals or missing persons, tracking retail visitor demographics and providing diagnosis of illnesses and conditions in healthcare. Meanwhile, technology firms are developing emotion detecting facial recognition applications, reportedly creating a new $20 billion industry.
According to a study by Boston’s Northeastern University, people make the expected facial expression about 20 to 30 percent of the time. For example, smiling when happy. However, although these expressions often go unnoticed, the rest of the time people’s faces move in specific ways according to the situation. Emotional AI systems are being optimised and trained to recognise happiness, sadness, disgust, fear, anger, surprise and other emotions.
Amazon Web Services (AWS) announced this week that its facial recognition software, Rekognition, can detect a range of emotions including a person’s fear. The software has already been made available to developers as an AWS cloud service.
Again, there are many applications for emotion recognition. Security authorities are interested in this technology to use as a tool to spot unusual behaviour in high security areas such as airports. Hospitals are looking at emotion detecting AI to help assess the level of pain or discomfort felt by patients and even diagnose conditions. Then, of course, the world of commerce is already trialing emotion recognition systems, in order to evaluate consumer reactions to advertising, promotions, products and services.
Emotion recognition doesn’t always need your face scan to start assessing how you are feeling. Software developers and behavioural scientists are developing systems that can help judge someone’s mood by the way that move, such as the way they approach a product in a shop or by tracking other signals from body language. Startups are already working with some of the world’s biggest brands on using emotional AI to inform their advertising campaigns.
The increasing spread of emotional AI, brings with it new concerns about privacy, personal data rights and, well, freedom of expression (although, in a sense perhaps, not thought about much in the past). The data that the emotion recognition captures could be considered to be biometric data, just like a scan of your fingerprint. However, there is little legislation globally that specifically addresses the capture, usage rights and privacy considerations of biometric data. Until GDPR.
The General Data Protection Regulation was introduced as regulation under EU law in May 2018. GDPR was developed to be comprehensive, covering data protection and privacy for all individual citizens of EU and EEA (European Economic Area), whether they are resident in the EU or not. This regulation does cover biometric data.
GDPR provides regulation for more than half a billion EU citizens, protecting an individual’s right to be ‘forgotten’, and requiring clear and affirmative consent from individuals before data is captured. Unlike its counterparts and predecessors, GDPR regulates biometric data to include the processing and storage of any personal data resulting from ‘the physical, physiological or behavioural characteristics of a natural person’.
So, although this has yet to be tested, it seems that the capture and processing of the emotions of EU citizens by new emotional AI without their explicit consent is prohibited by GDPR.
Well, isn’t that something!
This feature was first published by Carrington Malin on Linkedin.
Carrington Malin is an entrepreneur, marketing professional and advisor who has worked across almost every sector of technology. He helps companies, startup ventures and public sector organisations develop marketing strategies, digital initiatives and leverage new marketing technologies. He also publishes a daily Asia AI News digest. You can connect with Carrington on Twitter @CarringtonMalin or via Linkedin https://lnkd.in/furZ3s9