Technology is changing our lives in some of the most significant ways.
And while it might sound like something out of a sci-fi movie, a new Artificial Intelligence (AI) model that has been tested across eight cities in the United States claims to be able to predict crime a week ahead of time, with a high degree of accuracy.
Developed by a Chicago University Professor, the model was tested in Chicago and seven other cities in the US, boasting it can predict future crime with 80 to 90 percent accuracy.
Predicting crime trends
The software crunched the statistics on crime in Chicago from 2014 to 2016, predicting the likelihood of certain crimes that would be committed the following weeks. The predictions were about both the types of crime that would be committed as well as the suburbs where they would occur.
The mode is reported to have had similar results across all cities, predicting crime in 8 to 9 out of ten times within a two-block geographical radius.
The developers made the data and algorithm publicly available to increase scrutiny and transparency – because previous technology driven crime predictors used in Chicago had been found to have a racial bias.
While there is still some way to go, the developers believe that the tool has a role to play in assisting to develop high level social and law enforcement policy, along with supporting the allocation of funding and resourcing, rather than as a reactive tool for police.
The developers say the technology demonstrates ‘the importance of discovering city-specific patterns for the prediction of reported crime, which is then able to generate a fresh perspective on neighborhoods in the city, enabling prior investigation and questioning and also the opportunity to evaluate police action in new ways’.
The criminal justice system, like most other industries, has been exploring the role AI can play in improving outcomes for a number of years.
Australia without lawyers
Here in Australia, a ‘firm without lawyers’ was launched in the Northern Territory in 2017 which uses software called Ailira to assist clients with many issues relating to estate planning and establishing various other structures such as companies and partnerships.
Because there’s no actual lawyer involved, the firm says it can save clients a heap on legal costs.
Use of technology by government and law enforcement
Police and Governments in Australia are already using AI facial recognition technology. Across New South Wales high-definition mobile cameras have been in place for the past three years. They are capable of catching drivers at night, in poor weather conditions and at speeds of up to 300km/h.
In the immediate weeks after the cameras were deployed concerns were raised about the fact that images would be reviewed by Artificial Intelligence, and that the automated system could result in the issuance of fines to those who have done nothing wrong – in which case the onus would be unfairly placed on motorists to contest the fines. At the time, authorities put in place processes whereby all images are checked by a real human prior to fines being issued so that innocent people don’t end up being flagged as offenders.
Around the world, and here too, AI is also successfully being used in crime-solving – AI in laboratories can investigators with DNA testing and other analysis, particularly where the samples are old or low-grade. This can mean that historic crimes are able to be solved. AI can also be used to analyse case-law data and
Another application of AI is predictive justice, which is the statistical analysis of a large amount of case law data – to help Magistrates and Judges to view past court decisions if necessary when they are evaluating current sentencing options. AI can also predict recidivism by analysing criminal justice-related data, enabling more focused allocation of resources.
Like everything else, AI technology has its pros and cons.
Most experts agree that it can make a positive contribution, although they also warn that it is still an evolving technology, and the most important consideration with regard t the law is that its use is legislated so that AI is applied ethically – with human rights (such as privacy and non-discrimination) a primary consideration, and to reflect that it is a tool only, not a complete replacement for human skills, knowledge and experience.