A naukri.com initiative
Medium
2d
237
Image Credit: Medium
Your local police station might already be using AI to monitor your every move.
- Police departments across America have begun deploying AI surveillance cameras which have the ability to judge and categorize subjects.
- One such company selling the tech is Fusus, which is providing AI surveillance models to police stations across the country.
- Facial recognition algorithms often used in the technology, however, are fundamentally biased because of variations in the data used to train them.
- These disparities in data accuracy could often lead to the identification of innocent people.
- AI-powered facial recognition algorithms are exacerbating existing racial and class biases in policing, making it difficult to remedy the situation.
- Centralized systems such as these can identify clothing, vehicles, and individuals from various camera feeds making it easy to surveil a person's movements.
- These systems allow law enforcement to expand surveillance beyond the humanly possible, exacerbating existing biases prevalent in the police department.
- The datasets used to train these algorithms are often overwhelmingly composed of lighter-skinned subjects, which is causing biases to be embedded into the algorithms from the start.
- AI technology is marketed as more reliable than human decision-making, but instead of eliminating bias, AI is entrenching biases even deeper into the system.
- In essence, these technologies are not just about protection, they're tools of control.
Read Full Article
14 Likes
For uninterrupted reading, download the app