In September 2020, DHS proposed expanding the biometric data it collected from applicants trying to enter the country and from its crackdown on undocumented immigrants. The proposal drew fierce objections from privacy advocates.
Some have raised concerns that the move by ICE may be “a gross expansion of government surveillance threatening privacy, civil liberties, and people of color in particular without demonstrative security benefits”.
As late as October 2020 – the month before the US elections – DHS continued to purchase geolocation mobile phone data to carry out its missions. By cross-referencing property records and other data sets with mobile phone location data, investigators are able to track every move of individuals – including US citizens –without a warrant. This ability for live location tracking provides an unchecked avenue for extensive surveillance.
With the pandemic spurring some US cities to roll out contact-tracing apps and cameras capable of facial recognition, mask detection and temperature scanning, many are questioning the reach of such technology, its place in a post-pandemic world and how it might be used after the current public health crisis.
The strongest signal of what to expect from the Biden administration is a White House memo, which makes clear that while the construction of a physical wall would cease, renewed efforts to “supplement existing border resources with technology and infrastructure” would still be pursued.
Privatisation without privacy
Given its emphasis on ‘smart’ borders “powered by biometric data, artificial intelligence, facial recognition, aerial drones, infrared cameras, motion sensors and radar”, the new administration may choose to continue the work done by DHS. The department released its AI strategy right at the end of President Trump’s tenure, declaring itself a leader in AI innovation.
The strategy called for “smart engineering” and more transparency by presenting data documentation to both operators and affected populations outside the typical governmental and military access pool.
The document also announced the creation of an AI advisory council that would “consider legal, compliance, classification, civil rights and civil liberties, and privacy implications and responsibilities related to DHS AI projects, methods and capabilities” and would uphold the ethical integrity of its AI systems.
But when those AI systems depend on private contractors such as Palantir, Google and Amazon, can we expect such a commitment to ethical integrity?
The lack of data privacy legislation in the US means that private companies upon which DHS has come to depend may determine for themselves the standards by which they abide. Some companies, including Google and Facebook, have created their own AI ethics boards. These provide internal oversight for the company’s decisions, but often fall short: marking your own homework is never a good idea.
Biden’s ‘smart border’ approach is dangerous: it relies on the intrusive, data-driven approach of the former president in the name of a more humane approach to immigration. But the renewed emphasis on AI forces the US on to a surveillance path that has the potential to be far worse than any physical barrier, especially without clear regulation on private entities’ use of AI.
One thing is clear, however: President Biden is fully embracing AI technology for border security. Privacy and migrant advocates should pay the same – if not closer – attention to how DHS leverages commercial and governmental big data in the future, as they did over the last four years.
