Code of Ethics
Our Commitment:
AI technology that respects privacy and works for all races, ages, and genders
ROC believes in a just, non-violent world of equality and fairness. We prize democratic values, civil liberties and open and informed debate. When used to further these values, automated face recognition can continue to make the world a safer, better place for everyone. We have developed this code of ethics (“Code of Ethics”) to set forth the principles that we believe should guide any development and use of face recognition technology. This Code of Ethics establishes important limitations that we believe are appropriate in how face recognition should be utilized, including by public authorities and law enforcement and in legal proceedings.
This Code of Ethics serves as a guideline both for how we will strive to develop face recognition systems and how we will expect our integration partners, customers and end-users to develop and utilize face recognition systems based on our algorithms. We will require all of our partners and customers who use ROC face recognition algorithms to commit to abide by this Code of Ethics or develop a policy implementation guide specific to their custom deployment of the technology that assures that freedom of speech, freedom of assembly and the other civil liberties and rights recognized under the U.S. Constitution are protected and preserved.
First Principle
Face recognition should be used to make the world safer, more secure and more convenient while minimizing harm through proper workflows and controls designed to identify and mitigate sources of error and bias.
Private Commercial Use & Non-Law Enforcement Uses by Public Authorities
Face recognition use by commercial entities should require notice and consent, except when used for physical security purposes.
Similarly, face recognition use by public authorities should require notice and consent, except when used for national security, law enforcement or physical security purposes.
Face recognition use must be in compliance with any laws and regulations relating to privacy, data security, data protection, and processing, transfer, disclosure, sharing, storing, security and use of personal information, as applicable in any jurisdiction where it is used.
Law Enforcement and Legal Proceedings
Face recognition should not be used to perform real-time mass video surveillance against large databases of law-abiding U.S. citizens.
Before adding a person to a face recognition watchlist for any targeted real-time video surveillance, there should be reasonable suspicion that such individual is a perpetrator, victim or witness of a crime.
Video surveillance should never be used to suppress freedom of speech, freedom of assembly or any other civil liberty or right recognized under the U.S. Constitution.
Face recognition should not be used as the sole support of probable cause for arrest, search or seizure of any U.S. citizen or any property. Independent evidence should be required to establish probable cause.
Face recognition to establish an investigative lead should utilize best practices and workflows established by the Facial Identification Scientific Working Group (FISWG) and the Organization of Scientific Area Committees for Forensic Science (OSAC) Facial Identification Subcommittee, which require a trained human facial examiner to make final determinations based on morphological matching guidelines.
The fact that face recognition was used, the raw image or video data inputs and any human expert’s determinations should be discoverable facts in criminal proceedings.
Face recognition use must be in compliance with public authority and/or law enforcement policies and procedures, all ordinances, statutes and regulations, applicable court orders and supervisory frameworks and all U.S. Constitutional limits that protect civil liberties and rights.