Technology News
All About technology

Technology is no panacea for custodial deaths

0 4


It can never be an alternative to compassionate policing based on trust between the police and citizens

It can never be an alternative to compassionate policing based on trust between the police and citizens

India has a grim record in police brutality and custodial violence. Between 2001 and 2018, 1,727 persons died in police custody, but only 26 policemen were convicted for such deaths. The recent spate of custodial deaths in Tamil Nadu has yet again highlighted the methods used by the police during interrogation. It is not uncommon knowledge that the police, when they grow increasingly frustrated with the trajectory of their interrogation, sometimes resort to torture and violence which could lead to the death of the suspect. Custodial deaths are common despite enormous time and money being spent on training police personnel to embrace scientific methods of investigation. This is because police personnel are humans from different backgrounds and with different perspectives.

Use of technology

Given the problem of custodial deaths, technology has been proposed as a silver bullet by many. Several technological solutions are available to help prevent custodial deaths. These include body cameras and automated external defibrillators. There is no doubt that technology can help avert police custodial deaths. For example, body cameras could hold officers liable. Deception detection tests (DDTs), which deploy technologies such as polygraph, narco-analysis and brain mapping, could be valuable in learning information that is known only to a criminal regarding a crime.

Among the DDTs, the Brain Fingerprinting System (BFS) is an innovative technology that several police forces contemplate adding to their investigative tools. BFS has proved helpful for solving crimes, identifying perpetrators, and exonerating innocent suspects. Laboratory and field tests for the BFS at the Federal Bureau of Investigation, the Central Intelligence Agency, and U.S. Navy demonstrated no errors and no false positives and false negatives. The technique helps investigative agencies uncover clues in complicated cases. In June 2008, India convicted an accused leaning on evidence from a BFS device. In 2010, the Supreme Court, in Selvi v. State of Karnataka, rendered the evidence inadmissible. The court observed that the state could not perform narco analysis, polygraph, and brain-mapping tests on any individual without their consent. With informed consent, however, any information or material discovered during the BFS tests can be part of the evidence. As the BFS is high-end technology, it is expensive and unavailable in several States.

Police departments are increasingly using robots for surveillance and bomb detection. Many departments now want robotic interrogators for interrogating suspects. Many experts today believe that robots can meet or exceed the capabilities of the human interrogator, partially because humans are inclined to respond to robots in ways that they do to humans. From his studies, human-computer interaction (HCI) researcher Joseph Weizenbaum concluded that suspects might be more receptive to opening up to automated conversational counterparts than the police.

Robots equipped with AI and sensor technology can build a rapport with the suspects, utilise persuasive techniques like flattery, shame and coercion, and strategically use body language. Researchers at the University of Arizona have created automated interrogation technology called The Automated Virtual Agent for Truth Assessments in Real-Time (AVATAR). The Canadian Border Services Agency tested AVATAR last year. The HCI system uses visual, auditory, near-infrared and other sensors to scrutinise a suspect’s eye movements, voice, and other qualities throughout an interaction. The aggregation of information and its analysis by the system have been highly accurate.

Artificial Intelligence (AI) and Machine Learning (ML) are emerging as tool of interrogations. AI can detect human emotions and predict behaviour. Therefore, these are also options. ML can in real-time alert superiors when police are meting out inhumane treatment to suspects.

Valid concerns

There is a lot of concern about AI or robot interrogations, both legally and ethically. There exists the risk of bias, the peril of automated interrogation tactics, the threat of ML algorithms targeting individuals and communities, and the hazard of its misuse for surveillance. Therefore, while the technology available to the police and law-enforcement agencies is constantly improving, it is a restricted tool that can’t eradicate custodial deaths. While it might provide comfort and transparency, it can never address the underlying issues that lead to these situations.

What we need is the formulation of a multi-pronged strategy by the decision-makers encompassing legal enactments, technology, accountability, training and community relations. The Law Commission of India’s proposition in 2003 to change the Evidence Act to place the onus of proof on the police for not having tortured suspects is important in this regard. Besides, stringent action must be taken against personnel who breach the commandments issued by the apex court in D.K. Basu v. State of West Bengal (1997). The draft bill on the Prevention of Torture, 2017, which has not seen the day, needs to be revived. Technology may make policing more convenient, but it can never be an alternative for compassionate policing established on trust between the police and the citizens.

K. Jayanth Murali is an Indian Police Service officer

Leave A Reply

Your email address will not be published.