Even if we don't always acknowledge it, artificial intelligence (AI) has ingrained itself so deeply into our daily lives that it's difficult to resist.
While ChatGPT and the use of algorithms in social media have received a lot of attention, law is a crucial area where AI has the potential to make a difference.
Even though it may seem far-fetched, we must now seriously examine the possibility of AI determining guilt in courtroom procedures.
The reason for this is that it calls into question whether using AI in trials can still be done fairly. To control the use of AI in criminal law, the EU has passed legislation.
There are already algorithms in use in North America that facilitate fair trials. The Pre-Trial Risk Assessment Instrument (PTRA), the Public Safety Assessment (PSA), and Compas are a few of these. The employment of AI technology in the UK criminal justice system was examined in a study produced by the House of Lords in November 2022.
Empowering algorithms
On the one hand, it would be intriguing to observe how AI can greatly improve justice over time, for example by lowering the cost of court services or conducting judicial proceedings for small violations. AI systems are subject to strict restrictions and can avoid common psychological pitfalls. Some may even argue that they are more impartial than human judges.
Algorithms can also produce data that can be used by lawyers to find case law precedents, streamline legal processes, and assist judges.
On the other hand, routine automated judgements made by algorithms might result in a lack of originality in legal interpretation, which might impede or halt the advancement of the legal system.
The artificial intelligence (AI) technologies created for use in a trial must adhere to a variety of European law documents that outline requirements for upholding human rights. Among them are the Procedural European Commission for the Efficiency of Justice, the 2018 European Ethical Charter on the use of Artificial Intelligence in Judicial Systems and their Environment, and other laws passed in previous years to create an effective framework on the use and limitations of AI in criminal justice. We also need effective supervision tools, though, including committees and human judges.
Controlling and regulating AI is difficult and involves many different legal areas, including labour law, consumer protection law, competition law, and data protection legislation. The General Data Protection Regulation, which includes the fundamental principle for justice and accountability, for instance, directly applies to choices made by machines.
The GDPR has rules to stop people from being subject to decisions made entirely by machines with no human input. This principle has also been discussed in other legal disciplines.
The problem is already here; in the US, "risk-assessment" technologies are used to support pre-trial determinations of whether a defendant should be freed on bond or detained pending trial.
Sociocultural reforms in mind?
Given that law is a human science, it is important that AI technologies support judges and solicitors rather than taking their place. Justice follows the division of powers, exactly like in contemporary democracies. This is the guiding principle that establishes a distinct division between the legislative branch, which creates laws, and the judicial branch, which consists of the system of courts. This is intended to defend against tyranny and protect civil freedoms.
By questioning human laws and the decision-making process, the use of AI in courtroom decisions may upend the balance of power between the legislative and the judiciary. As a result, AI might cause a shift in our values.
Additionally, as all forms of personal data may be used to predict, analyse, and affect human behaviour, using AI may redefine what is and is not acceptable activity, sometimes without any nuance.
Also simple to envision is the evolution of AI into a collective intelligence. In the world of robotics, collective AI has silently emerged. In order to fly in formation, drones, for instance, can communicate with one another. In the future, we might envision an increasing number of machines interacting with one another to carry out various jobs.
The development of an algorithm for fair justice may indicate that we value an algorithm's abilities above those of a human judge. We could even be willing to put our own lives in this tool's hands. Maybe one day we'll develop into a civilization like that shown in Isaac Asimov's science fiction book series The Robot Cycle, where robots have intelligence on par with people and manage many facets of society.
Many individuals are afraid of a world where important decisions are left up to new technology, maybe because they think that it might take away what truly makes us human. However, AI also has the potential to be a strong tool for improving our daily lives.
Intelligence is not a state of perfection or flawless rationality in human reasoning. For instance, mistakes play a significant part in human activity. They enable us to advance in the direction of real solutions that advance our work. It would be prudent to keep using human thinking to control AI if we want to expand its application in our daily lives.
