Will lawyers be replaced by robots?
Artificial intelligence is not only around us but it is part of us
Better, faster, more productive... these used to be words to describe what we wanted from our internet broadband connection.
Now they describe what we expect not only from all our products and services but also from each other. The expectations that we have from the world and the world has from us are dynamic, ever-changing and transforming at speed.
By combining a basket of different technologies, artificial intelligence (AI) is increasingly seen as the answer to these unprecedented expectations and needs:
- Siri, on your iPhone, is artificial intelligence
- Anti-lock brakes in your car, learning the road conditions is artificial intelligence
- Amazon's recommendations on things that you might like is artificial intelligence
AI is all around us. Complicated and repetitive tasks are better performed by computer algorithms and mechanisms than by the human mind. The margin of error is narrower, the work is done more quickly and the training required is minimal.
Around the world the justice and legal systems have started to embrace the use of technology and the benefits of artificial intelligence. Data analytics are used, for example, to help predict crime and shape social policy.
Firms all around the world have improved the way they conduct legal research by using systems like ROSS. Developed by IBM, it is a tool which allows legal practitioners to use natural language and ask questions, rather than use keywords. ROSS then provides citations and suggests topical readings from a variety of sources.
The uses of artificial intelligence do not stop there, however. I recently looked at an app that:
- tracks how many steps we take
- tracks how many hours of sleep we have, and even
- tracks our heartbeats.
We are living in a world in which we have become part of the app.
Our phones and gadgets seem to know us better than we know ourselves. They can even pass judgement on our lifestyle choices, such as congratulating us when we hit a certain fitness target. Mine has broken, as it seems to have forgotten to congratulate me...
We must face the fact that artificial intelligence will become more entrenched in our lives. There is a correlative duty for us as solicitors to scan the horizon, think of the consequences of its uses and, where needed, we must act.
What are the consequences?
There is a wealth of research on the ethical consequences of artificial intelligence. However, I just want to focus on the legal effects that result from this new use of technology.
Use of data
Artificial intelligence relies heavily on the use of personal and corporate data for all its practical applications. This inevitably raises serious issues of privacy and data protection. How will 'big data' - such as your search engine history, your online banking, your medical history - be collected, be used, and stored? For what purposes can that data be accessed and by whom? Who is responsible for keeping it safe? How will we handle global data breaches across different jurisdictions?
There is some domestic and emerging international legislation on data protection, however, it is limited and reactive. The legal profession has a responsibility to start answering these questions now and to work together with policy makers and legislators to build a robust framework with appropriate safeguards.
Tort and accountability
Corporate technology giants, software developers and computer programmers are leading on innovation and setting the pace of a new world order.
Google estimates that in less than five years, there will be driverless cars on the streets of our main cities. This activity, which inevitably involves risk, raises questions of tort and accountability when things go wrong.
These self-driving systems may need to make split-second decisions that raise legal questions. A child suddenly runs into the road and the car has to choose: hit the child or swerve into an oncoming bus.
- How does the car decide?
- Who decides what the car decides?
- Who is liable if it makes the wrong decision?
Legal practitioners and academics need to come together to think ahead of the consequences of the new uses of technology and put forward ideas for the best legal response.
Changes to traditional areas of the law
Artificial intelligence is increasingly pushing the boundaries of the traditional areas of law.
The introduction of 3D printers has posed some serious legal questions. Although the technology used by 3D printers is not strictly AI, there is emerging thinking on the convergence of 3D printing, nanotechnology and even neurotechnology which is already making machines more intelligent.
There is now the possibility of emailing the design of an object to print it on the other side of the world. Then a 3D printer is used to produce this object. In the UK, the Royal Mail launched a 3D printing system in 2014.
This new model raises a number of questions:
- Who is the owner of the object?
- Who is accountable if the product does not fulfil its specifications?
- Who is responsible if the product is unsafe?
This example shows that there are new ways of ownership that are not currently predicted by our country's statutes or case law. Consequentially, it may generate new and previously unthought-of economic models.
The issue becomes even more problematic when technology is used for a sinister purpose. Criminal enterprises are known for using technology to further their aims.
- What will be the consequences of using a gun that was developed and produced using a 3D printer?
- How will the judiciary determine the type of offence? Who will be charged?
- What legal measures are needed to prevent these incidents?
It is our responsibility, our duty, to map such new uses of technology against the scenarios of risk in order to enable us to plan effectively for the future.
Role of law societies and bar associations
The majority of the questions which I have raised do not currently have answers in the law. Artificial intelligence is dictating the way that we do the law. We, as a profession, seem to be playing catch-up.
Corporate technology giants, programmers and software engineers have already imagined the future and are executing their vision.
Returning to the original question: will robots replace solicitors?
Many of our clients are already using the web for self-diagnosis. Many of them will Google their legal problems before they come to see us. Self-diagnosis, facilitated by artificial intelligence, is not and will never be a complete replacement for solicitors - we do so much more than dispensing black-letter law.
An artificial intelligence system is designed to stimulate human thinking but not creative or independent thought. Both of these qualities are essential for the legal profession and the discharge of our professional legal obligations of upholding the rule of law and the proper administration of justice.
Artificial intelligence is incapable of developing creative legal arguments that are needed in both contentious and non-contentious matters. Its ability of interpreting data is also extremely limited.
Although 'robots' might come up with a shortlist of relevant precedents, statues and regulation, they lack the ability to make a persuasive argument that takes the context into account, the individual circumstances of a client and, most importantly, the human experience.
No algorithm exists to replace us and the work that we do. Just because an algorithm might perform efficiently it does not mean that it will be correct. People do not want to be judged by an algorithm.
Finally, empathy and building trust with a client is an essential quality for solicitors which artificial intelligence systems lack. We have moral and ethical responsibilities in the work that we do, whereas AI does not.
Solicitors are and should, however, continue to use artificial intelligence to improve their work - for example, in digitisation of documents or legal research. These drive efficiencies, increase accuracy, assist with client retention and have the potential to reduce overheads and drive greater profits.
Sophisticated legal research tools, such as ROSS, are very useful and even have parameters to make decisions, but they do not make the decisions themselves. Expert legal intervention through solicitors is always needed.
Artificial intelligence systems are still unable to think of a vision of a better future, to plan and execute its delivery accordingly. This is why collaboration between solicitors and 'robots' seems to be the way forward. We have already seen this in the medical profession, where sophisticated software tools are already being used for early diagnosis of cancer in primary care, helping doctors to identify symptoms more quickly and accurately.
The future of the legal profession is likely to follow suit.
A version of this article was previously published as a speech delivered by Jonathan Smithers at the UIA Conference.