Personal Attention.
Aggressive Defense.

Photo of Thomas C. Mooney

Why facial recognition is not foolproof evidence

On Behalf of | Jul 26, 2021 | blog, Criminal Defense

Technology and, more specifically, our dependence on technology has entered dangerous territory. Countless daily activities are governed solely through technological indicators. Everything from the morning wakeup alarm, our lunch order placed via mobile app and the traffic signals encountered during our day are essentially impersonal. Most citizens in Maryland never give their reliance on automated functions a second thought until trouble arises.

Facial recognition analysis

Many mobile phone users rely on their device’s ability to correctly identify their facial features as a security measure. Likewise, mobile apps that contain potentially sensitive information, such as financial records, also offer passwords dependent upon unique physical characteristic identifiers.

Although facial recognition platforms make daily life easier, they are not without complications. The probability of misidentification is nearly 5%, across all mobile platforms and devices. The likelihood of misidentification of individuals with specific features is even higher. Facial recognition of people of Asian or African descent is up to 100 times less accurate than with other races.

Admissibility as evidence

Despite the unrefuted inconsistencies with facial recognition applications, attorneys continue to battle this technology when it is presented as evidence. In many instances, there is no confirmed law or precedence that determines when facial recognition can be admissible in a criminal trial. Highly skilled criminal defense attorneys can effectively challenge the admissibility of specific evidence presented by the prosecution team.

Efficiency is key for virtually everyone in their personal and professional lives. Technology makes even the most complex tasks easier and faster. Convenience takes on new meaning. Despite the many benefits of conveniences like facial recognition, it would be a severe mistake to consider any automated technology foolproof or even honor those findings as material over a human eyewitness.