Introduction
Facial recognition solutions have gotten a bad reputation from the way the software was trained, how the solutions were deployed, the lack of governmental regulation on the burgeoning industry, and the unethical abuses of those utilizing the technology in the early years before attention was brought to it.
OLOID wanted to develop a usernameless and passwordless user authentication and access solution to help organizations improve efficiency, reduce costs, and increase profits, and biometrics like Palm and Facial Recognition were essential components. But first, they had to overcome the mistakes made by others before them in designing and implementing the technology.
Respecting Privacy
One of the primary ethical concerns about face-based recognition is violating people’s right to privacy. This technology collects and processes vast amounts of face-based data and often combines other biometric identifiers, and OLOID understands and advocates to its clients that it must only be done with explicit consent from individuals.
More than any other data breach, violation of face-based data concerns users of these systems, so it is essential that organizations obtain informed consent for face-based recognition technology from employees, vendors, and visitors to their facilities, so they are all aware that the technology is in use, their biometric data is being stored for the period of time needed for the system to function, and unreservedly consent to allow it as a condition of employment, doing business with or visiting the organization’s facilities.
Once an organization using OLOID’s solutions has gathered user’s biometric data, they must have robust measures in place to protect against potential misuse of this data such as identity theft and harassment.
If possible, companies can employ alternative authentication and access control solutions that do not employ facial recognition, as a concession to those who have strenuous objections, such as on religious grounds. OLOID’s solutions include several such alternative methods, including access cards, NFC (Near Field Communication), PINs, and QR codes.
Avoiding Mass Surveillance
No one wants face-based recognition technology to lead to a workplace that feels like a surveillance state, where individual freedoms are compromised, and democratic rights infringed upon. Both employers and employees want the technology to bring convenience and enhanced security without feeling under constant surveillance.
OLOID understands that balancing face-based recognition technology and individual privacy is crucial, and safeguards must be in place to prevent misuse. Clear guidelines and oversight mechanisms are needed to help balance technology’s benefits and protect individual rights.
OLOID’s system does not provide a constant surveillance condition, it instead checks for facial recognition from tablets that are either portable and shared between users as they are needed or are affixed at entry and exit control points such as to a plant, sensitive areas within it, or at workstations to ensure only authorized personnel access an area, equipment, or data.
Avoiding Racial & Gender Bias
Ethical considerations must be at the forefront of any design or implementation of facial recognition technology. The potential benefits are significant, but organizations cannot ignore the risks and unintended consequences.
Face-based recognition algorithms have been found to exhibit significant disparities in accuracy across different racial and ethnic groups. The use of face-based recognition systems can amplify racial and gender bias and thus can harm people’s work lives.
Early facial recognition systems showed more errors when identifying individuals with darker skin pigmentation. Additionally, gender biases had also been observed, especially with misclassifications related to non-binary or gender non-conforming appearances.
In a workplace, these biases can lead to inaccurate and problematic outcomes, such as wrongfully identifying people as someone else or not recognizing them at all and denying them access, leading to unauthorized access or production delays.
Face-based recognition can address bias through diverse and representative training data, algorithmic improvements, and ongoing testing to reduce discriminatory outcomes.
To avoid these problems at companies with very ethnically diverse workforces, often utilizing some level of worn PPE (Personal Protective Equipment), OLOID has developed its facial recognition software with these considerations and best practices in mind, and has achieved highly accurate results with ~0% false positives, that can capture identities accurately in all light conditions, and with liveness detection so it cannot be spoofed with pictures or video playback.
Recognizing Minority Groups
Bias mitigation involves diverse and balanced training data, algorithmic adjustments, regular audits, and continuous monitoring to ensure fair and equitable outcomes for all demographic groups.
OLOID understands that most of its clients have very ethnically diverse workforces, and has applied these techniques to achieve its ~0% false positive rate for all types of employees, including those with different facial structures, skin tones, facial hair, hairstyles, tattoos, and piercings of people and cultures from all over the world.
Self-Regulation
In the absence of lagging government oversight through regulations of any technology in any industry, self-regulation that exceeds expectations is always the best course of action.
With OLOID’s guidance from countless deployments, organizations should develop and deploy comprehensive frameworks for everything from the language on their consent forms to how long they store biometric data, where, for how long, and how they assure timely data sanitization when that data is no longer needed.
When setting a high bar for themselves, organizations should recognize that the regulatory landscape for face-based recognition technology varies across regions. Some countries have regulations, while others lack comprehensive frameworks. Existing rules often focus on data protection and privacy, but organizations need to anticipate and address specific requirements in other regions they do business in such as the EU or California.
Being ahead of the curve and setting internal policies and standards more stringent than lagging government regulation can help avoid fines from government regulatory agencies or legal action from the American Civil Liberties Union. Much like many cybersecurity and physical security-related issues, an ounce of prevention is worth a pound of cure.
Employing Ethical Principles
Ethical frameworks, such as transparency, accountability, fairness, and respect for individual privacy, should guide face-based recognition systems.
Facial recognition is not just a technological advancement, it has ethical and social implications and is about the values we want to uphold as a society.
Communication, collaboration, and consensus among an organization’s leaders, employees, and in-house or external ethicists can establish guidelines that prioritize privacy, reduce biases, and protect individual rights as well as the company’s public image, brand, and revenue when facing a consuming public that is increasingly buying from organizations that match their ethical and social values.
Align With Government Legislation
Considering the impact of face-based recognition technology on society, comprehensive government legal interventions are inevitable and already foreshadowed by laws in the EU and California. Governments will move to protect individual rights and establish regulatory frameworks ensuring accountability and transparency. Legislative approaches will range from specific bans to oversight and accountability measures.
Organizations deploying facial recognition solutions should be anticipating these legislative measures for their home country and others they do business in where they may be held to that region’s standards. Collaborative efforts among governments, businesses, other stakeholders, and civil society can shape legislation that balances businesses’ needs and protection of individuals’ rights.
Conclusion
OLOID offers facial recognition and other passwordless authentication solutions that address privacy, transparency, and fairness concerns. OLOID applications mitigate biases and promote fair practices. Its technologies provide multi-mode, simple authentication systems that meet human needs.
If you want to explore in more detail how the OLOID’s facial recognition solution was designed with all of these considerations in mind to overcome authentication challenges for frontline workers in countless organizations, lead to a fast return on investment (ROI) when switching from inefficient and costly systems to passwordless authentication, and improve companies’ bottom line in the process, visit our website and Contact Us to speak with a specialist about how we’ve done so for Tyson Foods and many other companies across diverse industries over the years.
