Facial Recognition Tech Threatens Human Rights, Needs Oversight

WASHINGTON — A new two-part study from the Center for Strategic and International Studies, a D.C.-based national security think tank, has said that civilian and judicial oversight will be crucial to limiting the potential human rights abuses from government use of facial recognition technology.

This would need to include operational rules for deployment by law enforcement, clarification about the use of any evidence that comes from this technology in court, and clear rules about data sharing, the report said.

Progress in the development of facial recognition technology in the last decade, in which software is used to match images of faces to identify or to verify identity, has made it available to governments and companies around the globe, according to CSIS.

The powerful tech has raised numerous issues about fundamental rights like privacy, nondiscrimination, freedom of assembly, freedom of religion, and freedom of speech, according to the report. It also frequently lacks transparency in the way it is deployed by governments, a group of panelists said at a release for the report on Tuesday afternoon.

The technology has been used by border control and by law enforcement looking to identify protestors, among other uses, prompting reports and allegations of bias, repression, and a lack of transparency. Concern remains high for the potential of this technology for what Steve Crown, vice president and deputy general counsel at Microsoft Corporation, characterized as “dystopian” uses, during Tuesday’s release.

Those who have strong democratic traditions are at the lead of the conversation about how to ensure human rights in deploying these technologies, Crown said, but those lacking traditions are behind.

It is also not clear if governments deploying this technology understand even the basics about how it works, what its limitations may be, or how it may curtail fundamental human rights.

Grecia Macias Llanas, a lawyer from Mexico who is working on these issues, explained that in her experiences in Mexico, law enforcement has not known or not disclosed the basics of the algorithmic technology they use.

The disparate impact this technology can have has proven hard to convey to officials, she added.

The CSIS report is part of a series attempting to map the whole supply chain of this technology, as well as the risks of its deployment for human rights by both governments and corporations. 

Of particular interest to the panelists was the mapping of how the industry is structured. 

Amy Lehr, a senior associate with the Human Rights Initiative at CSIS and the lead author of the report, said that understanding the supply chain is important. 

She also highlighted two areas of possibility for regulating this technology in the future, including setting up regulatory bodies that can give guidance, much like the U.S. Security and Exchange Commission does, and also establishing norms around the deployment of the technology. 

The report clarified that the regulatory regime ought to set out the “impermissible uses” of facial recognition tech, as well as establishing a legal basis for its use, encouraging standards to avoid discrimination, and create a framework for private firms involved in the production of this technology to perform due diligence on human rights. 

This would include making remedies available to people affected by the way the tech is used and strong limits on the data collection and use. 

It would also need an enforcement mechanism, the report said, to enact national law and human rights law.

Both parts of the study were made available Tuesday. Its release event was sponsored by the U.S. State Department Bureau of Democracy, Human Rights, and Labor of the State Department.