iBorderCtrl smart lie detection system

iBorderCtrl is a 'smart lie detection system' intended to strengthen security at the borders of the European Union, and to accelerate border crossings. Funded by the European Commission under the Horizon 2020 framework to the tune of EUR 4.5 million, the system comprises hardware and software, including biometric verification, document authentication and risk assessment. 

IBorderCtrl also features an automated deception detection, or video-based lie detection module, that comprises an avatar of a border guard that asks travelers 13 questions about their personal backgrounds and travel plans, and uses AI to analyse 'non-verbal micro-gestures'. Human agents are notified if people are assessed as being dishonest.

System 🤖

Operator:
Developer: European Dynamics
Country: European Union
Sector: Govt - immigration
Purpose: Detect traveller lies
Technology: Behavioural analysis; Facial recognition; Emotion recognition
Issue: Accuracy/reliability; Bias/discrimination - race, ethnicity, gender, age, disability; Human/civil rights; Privacy
Transparency: Governance; Marketing; Legal

Risks and harms 🛑

iBorderCtrl has been widely critcised as inappropriate and is seen as likely to result in unfair decisions affecting people's freedom of movement and liberty, and to violate privacy. 

The European Commission was also accused of inadequate transparency, with Patrick Breyer MEP taking the organisation to court to find out how the system was developed and tested.

Legal, regulatory 👩🏼‍⚖️

Research, advocacy 🧮