In Summary

  • Swinburne researchers claim artificial intelligence would improve efficiency and fairness of sentencing
  • AI can process many variables autonomously to assess crime and deliver a sentence without prejudice

Sentences handed down by artificial intelligence would be fairer, more efficient, transparent and accurate than those of sitting judges, according to Swinburne researchers.

Dean of Swinburne Law School, Professor Dan Hunter, and Swinburne researcher Professor Mirko Bagaric say artificial intelligence (AI) could improve sentencing procedures by removing emotional bias and human error.

In a paper for the Criminal Law Journal, Professors Bagaric and Hunter argue that AI sentencing would better identify, sort and calibrate all the variables associated with sentencing, including criminal history, education, drug/alcohol use, emotional motivations and employment.

The pair argue that sentencing decisions are often influenced by more than 200 considerations, many of which are variables which have been established prior to court hearings.

Professor Bagaric says subconscious bias plays a large part in sentencing in which judges or magistrates hand down harder penalties to offenders of a particular race or background.

He says that AI has the potential to make sentencing fairer, as it is incapable of adopting any emotional bias or prejudice towards people of certain profiles.

“It would make sentencing more transparent and quicker,” says Professor Bagaric.

“It would also eliminate judicial subconscious bias in sentencing that results in people of certain profiles, such as indigenous offenders, being sentenced more harshly.”

It would make sentencing more transparent and quicker.” - Professor Mirko Bagaric, Swinburne University of Technology

Professor Bagaric also points out that AI would make sentencing more consistent as a computer would quickly formulate calculated penalties, as opposed to relying on human intuition.

 “Currently, sentencing is a discretionary process, which means it is not a checklisted, methodical approach,” explains Professor Bagaric.

“Instead, it is an intuitive process, this leads to patent inconsistency because different judges have different intuitions, thereby resulting in judges prone to harsh or soft sentencing.”

For their research, Professor Hunter and Professor Bagaric conducted an in-depth analysis on whether AI could improve sentencing procedures by autonomously processing these variables.

While Professor Hunter believes that AI has the potential to radically improve the judicial system, he says it is still far from being publicly accepted.

“We are a long way from community acceptance of machines passing sentence on humans, even if it is fairer and more just on the whole,” says Professor Hunter.

“People being judged by machines feels very Orwellian, or Terminator-like.”

He cautioned against society rushing in to replacing judges, but said AI could be used alongside judges, to allow for a better, more transparent, and ethical sentencing system.

“We will need to implement AI in a way that assists the judge in sentencing, rather than taking away all agency away from the judge,” he explains.

 “We need to demonstrate, through empirical research and extensive pilot programs, that this is a feasible, fair, and efficient way of sentencing that can help everyone.”

Professor Hunter and Professor Bagaric are working to develop research projects that will assess the effectiveness of using AI in conjunction with the traditional judicial process.

“Mirko and I are developing research projects to find the best way to ensure justice using technology,” explains Professor Hunter.

“That includes having courts and communities involved in the process, and ensuring the ethics of judging are complied with.”

Professor Bagaric’s and Professor Hunter’s article ‘Can Sentencing Be Enhanced by the Use of Artificial Intelligence?’ is published in the Criminal Law Journal.