When legal information becomes false confidence

AI is improving access to legal information, but it is also giving consumers the misguided confidence to commence proceedings they do not understand and cannot sustain.

AI is giving legal consumers access to legal information on a scale and at a speed we have never seen before.

This may substantially improve access to legal information, legal services and, ultimately, justice.

I say may.

It is also creating an urgent and practical problem for courts and tribunals.

People who previously had little or no access to lawyers can now obtain seemingly high-quality information, identify possible legal issues, and begin to understand their rights, often before, or even without, setting foot in a lawyer’s office.

VIEW ALL

That may be a good thing. Again, I say may.

The courts, however, are where the consequences of getting it wrong will ultimately be experienced. They must deal with cases brought before them regardless of how they were prepared or whether they should ever have been commenced.

This is not simply an issue for the legal profession, and especially not one for the judiciary alone to deal with and sort out.

It is a broader societal and whole-of-government issue with profound implications for how we manage disputes and dispute resolution.

Legal information mistaken for legal advice

AI gives people access to sophisticated legal tools presented in confident and reassuring language, often in a form like that used by a lawyer.

That is both a good and a bad thing.

This information can be useful. I stress the word information.

It can help people understand possible rights, identify causes of action, prepare questions, and evaluate a potential claim before seeking legal advice.

It is also problematic.

Lawyers are increasingly seeing clients arrive with AI-generated legal information that those clients understand to be definitive legal advice. That it came from AI seems only to add confidence rather than caution in many cases.

Legal consumers may therefore end up being more informed than ever before, or indeed much more misinformed.

The appearance of legal competence created by AI is not the same as the actual legal competence of a lawyer, particularly when the material is ultimately tested by a judge in a courtroom.

AI providers generally classify or disclaim what their systems produce as legal information. The consumer may nevertheless see something that is fast, free, and apparently better than any legal advice they have previously been able to access or afford.

They may believe it is all they need, without contemplating the value or cost of even asking a lawyer to check it.

That is the concerning part.

Experienced lawyers would rightly treat much of what AI produces as, at best, legal information. They would not treat it as legal advice or as a substitute for their own judgement and expertise.

The consumer may not appreciate that distinction when the material appears so attractive, convincing, and authoritative.

AI enables untrained consumers to produce professional-sounding pleadings, submissions, and correspondence with blind faith, false bravado, and misguided confidence in legal assumptions that justify their own take on a dispute.

The documents they produce, file, and rely upon may have little or no proper legal basis. In some cases, the prospects of success may be hopeless.

The danger is not merely that AI may produce something wrong. It is that it may present something wrong as certain, complete and authoritative.

That material may then be pleaded as black and white to courts accustomed to dealing with matters in shades of grey.

As I previously said to litigation clients and law students over the years – if it is such a black-and-white, open-and-shut case, why has it not settled?

While AI can generate plausible language convincingly, it cannot and should not take professional responsibility for whether the law is correct, whether the evidence supports the claim, whether the procedure is appropriate, or whether litigation is commercially and personally sensible.

It remains predictive technology at the end of the day. You still need to retain a lawyer for that, whether you like it or not.

The conundrum, of course, is that lawyers are expensive and, in many cases, unaffordable. Regrettably, the people who most need early, competent legal advice may be the people least able to afford it.

AI appears to fill that gap at the initial assessment stage, but without the judgement, responsibility, and accountability that proper legal advice requires of a lawyer.

That is the central tension we have now.

Intervention before filing

There is now a stronger argument than ever before for greater government investment in legal information, legal aid, and early legal and court triage.

This may have to include more lawyers, properly supervised law students, recent law graduates, and legal officers employed to triage matters at an early stage.

Their role would focus on legal process, legal information and procedural or quasi-legal assistance, rather than assuming responsibility for providing legal advice or conducting the case.

They could identify apparently defective or hopeless claims and refer them for timely legal review. Qualified lawyers would remain responsible for advising whether a claim has no hopeless prospects of success.

As I write the above, I appreciate how difficult, contentious, and costly this would be to implement. All this, however, cannot be left until proceedings have commenced and reached the post-filing case management stage.

By then, defective material may have been filed, positions may have hardened, and considerable financial and emotional investment may have been made by both sides in a claim that should not have proceeded.

Courts will nevertheless have to become involved at a much earlier stage through more intensive initial directions, identifying defective pleadings before substantial time and costs are incurred and referring suitable matters to early dispute resolution again much sooner than we have ever done before.

The court system, however, cannot become the lawyer for self-represented litigants. Courts are impartial and cannot assume responsibility for conducting a person’s case.

Basic procedural guidance and properly resourced early triage can, however, be provided without crossing that line.

Am I advocating more legal aid in the traditional sense? Possibly not.

Am I saying we may need a properly funded, court-connected but operationally independent pre-filing legal information and referral service? Possibly, yes.

Such a service would not decide who should win or give advice. It would help identify the dispute, explain the process, detect obvious legal or procedural problems, and direct the consumer to limited legal review before a claim is filed.

The alternative is overburdened courts, court officers and judges having to identify the actual legal issues in matters presented with the false confidence and false bravado of what self-represented litigants understand to be AI-provided legal advice.

Lawyers as officers of the court, and despite acting for the other side, may also be required to assist the court more than ever before, adding legal fees and costs for represented parties who may have done nothing wrong.

The represented party may ultimately win and still be left substantially out of pocket. The self-represented litigant may face serious financial and emotional consequences after pursuing and losing the case, including legal costs awarded against them.

The answer is not to discourage legal consumers from using AI. That will not assist, and it will not stop them.

The immediate task is to develop earlier assistance, better triage, and clearer responsibility without destroying the accessibility that makes the technology valuable.

Court-based rules and case management are necessary, but they deal with the problem once it has entered the legal system. The harder question is what society does before an AI-assisted dispute becomes filed litigation.

This is not merely a court-administration problem. It affects the cost of resolving disputes, the use of public resources, confidence in the justice system and the ability of ordinary people to make informed decisions about their rights.

AI is a significant opportunity for access to legal information, legal services and justice, but it remains both a sword and a shield.

The courts and legal profession will experience the consequences daily.

The issue is not whether courts should respond. They must. The issue is whether governments and society will provide the earlier assistance needed before courts are left to manage the consequences.

The dilemmas and their solutions belong to us as a society, not just the legal profession.

Paul Ippolito is a legal futurist and principal of Ippolito Advisory.