Proponents
Government supporters of data assessment tool often point to the high price tag of mass incarceration. In the United States, about eighty billion dollars a year are spent on incarceration. Many government officials believe that risk assessment tools are key to reducing incarceration rates. Some of the greatest examples of reducing incarceration revolve around cities and states reforming their bail system with data assessment tools. One Ohio county, which started using data assessment tools to compliment their bail system, was able to release an increased number of pre-trial detainees from jail without raising their pre-trial crime rate. The former Attorney General of New Jersey claimed similarly positive results throughout her state's justice system after New Jersey's law enforcement focused on data analytics. Supporters hope that relying on data-driven tools will help cities and states keep low-risk defendants out of jail, thereby reducing the prison system's financial burden nationwide.
While government officials care about budgets and money, they also have a deep interest in reducing violence and crime. Studies have shown that these tools, while imperfect, can competently predict violence. Likewise, officials have reported drops in crime in relation to the use of risk assessment tools [FT].
Artificial Intelligence has the potential to act as a more complex and efficient version of risk assessment tools in the near future. A self-correcting tool may be able to stimulate the release of more low-risk inmates and defendants while ensuring higher risk criminals stay behind bars. The epitome of efficiency in judicial decision-making would be for Artificial Intelligence to make the final determinations about the fate of defendants and prisoners.
Detractors
Some police officers have fought against the use of data assessment tools, arguing that they are insufficient to protect the public and allow recently caught criminals to go free. As of August 2017, New Jersey's January bail reform had already led to two homicides. Despite humans making the final decisions, some argue that judges are being asked to arbitrarily overrule or uphold scientific recommendations they do not understand. This inconsistency may be dangerous, leading to perilous shortcomings like the fatalities in New Jersey.
Allowing Artificial Intelligence to be the ultimate decision-maker in judicial decisions might satisfy some officials worried about capricious decision makers. After all, such a measure would streamline the judicial process, eliminating judges and the inconsistency that they bring to judicial decision-making. However, allowing Artificial Intelligence to decide criminal justice issues would create new government opponents. To become a judge, one generally must get an undergraduate and then law degree, pass the bar exam, practice as a lawyer for an extended period of time, and then be recognized as having such strong expertise, values and philosophy as to warrant being appointed or elected to a judgeship. Even then, many judges still must go through hours of government-mandated training. An Artificial Intelligence tool would just have to make predictions based on statistical models. Judges would likely find data analytics' singular approach incomplete, lacking the personal perspectives and experiences necessary to make holistic and just judicial decisions.
Government supporters of data assessment tool often point to the high price tag of mass incarceration. In the United States, about eighty billion dollars a year are spent on incarceration. Many government officials believe that risk assessment tools are key to reducing incarceration rates. Some of the greatest examples of reducing incarceration revolve around cities and states reforming their bail system with data assessment tools. One Ohio county, which started using data assessment tools to compliment their bail system, was able to release an increased number of pre-trial detainees from jail without raising their pre-trial crime rate. The former Attorney General of New Jersey claimed similarly positive results throughout her state's justice system after New Jersey's law enforcement focused on data analytics. Supporters hope that relying on data-driven tools will help cities and states keep low-risk defendants out of jail, thereby reducing the prison system's financial burden nationwide.
While government officials care about budgets and money, they also have a deep interest in reducing violence and crime. Studies have shown that these tools, while imperfect, can competently predict violence. Likewise, officials have reported drops in crime in relation to the use of risk assessment tools [FT].
Artificial Intelligence has the potential to act as a more complex and efficient version of risk assessment tools in the near future. A self-correcting tool may be able to stimulate the release of more low-risk inmates and defendants while ensuring higher risk criminals stay behind bars. The epitome of efficiency in judicial decision-making would be for Artificial Intelligence to make the final determinations about the fate of defendants and prisoners.
Detractors
Some police officers have fought against the use of data assessment tools, arguing that they are insufficient to protect the public and allow recently caught criminals to go free. As of August 2017, New Jersey's January bail reform had already led to two homicides. Despite humans making the final decisions, some argue that judges are being asked to arbitrarily overrule or uphold scientific recommendations they do not understand. This inconsistency may be dangerous, leading to perilous shortcomings like the fatalities in New Jersey.
Allowing Artificial Intelligence to be the ultimate decision-maker in judicial decisions might satisfy some officials worried about capricious decision makers. After all, such a measure would streamline the judicial process, eliminating judges and the inconsistency that they bring to judicial decision-making. However, allowing Artificial Intelligence to decide criminal justice issues would create new government opponents. To become a judge, one generally must get an undergraduate and then law degree, pass the bar exam, practice as a lawyer for an extended period of time, and then be recognized as having such strong expertise, values and philosophy as to warrant being appointed or elected to a judgeship. Even then, many judges still must go through hours of government-mandated training. An Artificial Intelligence tool would just have to make predictions based on statistical models. Judges would likely find data analytics' singular approach incomplete, lacking the personal perspectives and experiences necessary to make holistic and just judicial decisions.