Trial by AI – how far will artificial intelligence reshape our legal systems?
As the development of AI is on a seemingly unstoppable trajectory, we are increasingly being asked to consider a) how the technology will transform our lives and b) to what extent should we allow it? It is a debate on which there is little consensus, but which permeates almost every aspect of society. The law is not immune. Many in the legal community are beginning to question the role of AI within the industry and whether it could even replace lawyers and judges. At the very least, it could make the legal system more efficient and allow the judicial process to record more accurate and fair decisions.
Charles Kuhn is a partner and Neide Lemos is a trainee lawyer at Clyde & Co.
How can AI influence our legal systems?
On the face of it, there is certainly potential for AI to create a less biased legal system. Just as 94 to 96% of car accidents are caused by human error, the history of criminal justice is full of miscarriages of justice. The list is long – from the recent post office scandal in Britain and a wasted decade of Libor prosecutions in the US, to wrongful executions on death row, the Birmingham six and Andrew Malkinson. It's clear that the criminal justice system doesn't really work for everyone.
AI-led proceedings do not require juries. And if algorithms were continually updated and improved over time, human biases could be significantly reduced. Furthermore, tampering by juries or judges could be nearly eliminated. After all, it is difficult to interfere with or intimidate an AI system, or even buy it off.
And instead of banning a human-run jury from using the internet or investigating cases for fear of biased information gathering, an AI-powered system would collect all available data to inform its conclusions – making hung juries a thing of the past.
Some would suggest this goes too far. But even for those who prefer a human head (and heart) at the helm of legal proceedings, AI can certainly help increase efficiency. For example, when handling and recording data. Using AI to analyze large amounts of case material and verify information could reduce the time to trial for a large number of cases and ensure that they are heard much more quickly when they reach the courtroom. For example, it was used in a recent trial at the Old Bailey, which saw the murder case of six-year-old Rikki Neave reopened. AI software enabled review of more than 10,000 documents to identify patterns in the evidence that were missed during the initial inspection.
The impartiality of AI, even at an early stage of research, could spot patterns and solve the facts faster. And powered by machine learning and the right deep learning techniques, AI could even be applied to highly complex white-collar investigations. For example, insider trading cases could benefit from AI's ability to understand patterns and irregularities in various sources, such as monitoring financial transactions and flagging suspicious activity.
How is it already used?
Some countries around the world are already testing the technology in their legal systems, with significant success. China uses Xiao Zhi 3.0 ('Little Wisdom') to support legal decision-making in civil proceedings. It was first deployed in Hangzhou in 2019 and facilitated a trial against ten people who were unable to repay their bank loans. Now it is used in simple financial disputes, analyzing case materials, verifying information from databases and recording testimony. The implementation of Xiao Zhi 3.0 in the Chinese legal system has made it faster and more efficient; Ten separate trials can now be heard in one 30-minute hearing before one judge. Such support within the UK courts could be instrumental in addressing the significant backlogs that currently exist across the country.
In Germany, the “FRAUKE” program was tested for eight weeks at the Frankfurt court in passenger rights cases. It was successfully used to extract important elements from procedural observations and draft proposals for judicial decisions using similar cases. After the initial success, the Hessian Ministry of Justice approved further development of the technology. It is also not the first time that AI has been used in Germany: the AI, “OLGA”, has been used at the Oberlandesgericht Stuttgart in diesel cases, which is almost always appealed. OLGA analyzes information by automating cases into case groups using jury-determined factors (such as vehicle, engine type, mileage). This makes it possible to plan and handle matters more effectively.
Barriers to adoption
There is no doubt that digital innovation will play an important role in our legal systems – even the simplest programs can help establish and organize the facts of a case. But can we ever trust AI to take full control when it comes to dispensing justice? AI is not programmed to reach a binary conclusion; it learns from scenarios and applies its own conclusions. These scenarios are not predefined; the program just 'figures out' what to do. That's a lot of trust to place in the hands of technology.
Skeptics would argue that the country will never truly have 'common sense'. Of course it can generate human reactions. But it lacks people's intuitive understanding. Can AI experience empathy and a true sense of justice? To achieve something similar, deep data and neural networks are used – but there are many who question whether this is enough.
Meanwhile, advocates have suggested that AI should further deploy legal aid to read body language and decide on the veracity of witnesses. This is enough to alarm legal experts and psychologists around the world, who will argue that many factors influence body language and testimony, including stress and misleading questions. Furthermore, lawyers would worry that the application of this type of technology violates the principles of self-incrimination, which begs the question: should this type of analysis be considered at all?
Then there is the skills of our existing judges and senior lawyers. Many of these are more Luddite than technophile and are a stumbling block to initial implementation. For example, despite some interesting AI testing cases, German judges and prosecutors are strongly resisting any attempt to even digitize and record criminal proceedings. In Britain, many judges who have convicted Post Office submasters on the claim of digital 'robustness' are said to have been unable to write or verify computer code.
Taste by AI
At the current pace of AI development, it seems only a matter of when, not if, AI will become part of our legal systems. It may take longer to develop and integrate more complex AI applications for decision-making, while simpler AI tools for research and document review are already in use. What is clear is that AI's enhanced capabilities will leave ours in the dust, leading to greater efficiency, more accurate decision-making, and even less bias. But as progress continues, there are still important ethical and practical questions to be answered about the extent to which we allow technology to influence our legal processes.
In the US and other jurisdictions, defendants are asked whether they prefer to have their case tried before a jury or a single judge. It won't be long before they are actually tried by AI.
We've highlighted the best AI writers.