You have 0 free articles left this month.

Concerns notices cannot be ‘generated from a template’

As the cost of legal representation puts justice beyond the reach of many Australians, artificial intelligence offers an alluring shortcut. But access to justice is not simply a matter of making legal arguments easier to produce; it is a question of whether they are good enough to be heard. And, in defamation law, where the stakes are reputational and the nuances unforgiving, AI may be making the courthouse doors easier to approach without bringing litigants any closer to justice.

November 18, 2026 • By Jerome Doraisamy
Share this article on:
expand image

There is something paradoxical about AI’s arrival in the courtroom. A technology promising to make knowledge universally accessible is making the administration of justice considerably more difficult.

Australian courts are discovering that the ability to generate a legal argument is not the same as understanding one. ChatGPT can produce submissions, affidavits and elaborate accounts of alleged wrongdoing in minutes. Whether any of it has legal merit is another question.

 
 

The scale of the problem is becoming harder to ignore. As reported by The Australian Financial Review in October, self-represented claims in the Federal Court had increased by 80 per cent in a year. Chief Justice Debra Mortimer has described the expanding workload and errors associated with AI-assisted litigation. Meanwhile, judges and registrars face the unenviable task of distinguishing genuine grievances from procedural clutter.

The response is revealing. Courts are not simply closing their doors to technological change; they are experimenting with earlier scrutiny, tighter filing procedures and clearer expectations of litigants. The challenge is to preserve access to justice without allowing access to become an invitation to overwhelm it.

The irony is that while AI has made it easier than ever to sound like a lawyer, it has not made it easier to have a case.

As Bartier Perry partner Adam Cutri and associate Zach Bentley note, digital media has made it “spectacularly easy” to defame and be defamed. Social media, podcasts, and streaming, the pair told Lawyers Weekly, “have created an unprecedented pool of potential plaintiffs”.

“And, while the serious harm threshold was designed to keep the lightly bruised out of court, AI may now be weakening another gatekeeper: the concerns notice,” the pair said.

Danielle Snell, managing partner of Elit Lawyers by McGirr & Snell, said AI has the potential to improve access to justice by helping people understand their defamation and privacy rights and take the first steps towards vindicating their reputation. But, she added, “in defamation matters, a concerns notice is not simply a document that can be generated from a template”.

An uptick in claims?

A “regrettable impact” of the age of AI, O’Brien Criminal & Civil Solicitors special counsel Stewart O’Connell mused, is that some people feel they can now rely on it rather than sound legal advice from an experienced lawyer.

The dangers of this are particularly pronounced in defamation, he said.

“Defamation is a technically complex and nuanced area of law that at its core relies on the meaning of words as understood by an ordinary human. As smart as AI is, it can never understand the meanings that an ordinary human would give to words, particularly given that this always changes depending on context (including slang),” he said.

Thomsons senior associate Joel Parsons pointed out that while AI has the potential to improve access to justice and reduce legal costs, at the same time, there’s a risk that the uncritical use of AI leads to ill-advised or underdeveloped complaints. “That’s a risk in defamation, as in other areas of law, and we are seeing increased use of AI to generate concerns notices. It’s possible that will translate into more defamation proceedings before the courts,” he said.

This aligns with the observations of Parsons’ colleague at Thomsons, partner Marlia Saunders, who said that given lawyers have professional obligations to sign off on court pleadings as having reasonable prospects of success, whereas self-represented litigants using AI do not, “more unmeritorious claims are being commenced”.

“Since these litigants are not receiving proper legal advice based on practice and experience, they can also have inflated expectations of the veracity of their claim and the amount of damages or other remedies they want to receive,” she said.

Ripple effects from emerging technologies

So, what are defamation lawyers seeing on the ground right now?

BlackBay Lawyers managing partner Victoria-Jane Otavski said that a “large portion” of the concerns notices her firm has been asked to review and respond to recently have “clearly been put together with the help of AI”.

“You can usually tell from the length alone, because they run considerably longer than anything we were receiving a few years ago,” she said.

“To be fair, some of them feature a modest improvement, and parts even appear to meet the requirements of section 12A of the Defamation Act 2005 (NSW), but more often than not, they still miss the mark, and critically when addressing the serious harm alleged to have been caused by a publication.”

That is hardly surprising, she said, when the person writing the prompt starts from their own sense of grievance and the AI tool they are using is built to produce something they will be pleased with, rather than to tell them their claim won’t stack up.

The result, Otavski said, is that “aggrieved parties feel emboldened to press on into proceedings”.

“We have watched the same thing happen in employment and commercial disputes, and the courts are feeling it,” she said. “The best thing practitioners can do is engage early and say plainly where a notice falls short, and not to match volume with volume.”

That aggrieved parties can now produce a plausible notice in minutes, using AI, is good news for those with genuine grievances, with limited means, who are seeking access to justice, Cutri and Bentley suggested.

Afterall, the pair said, “protecting your reputation should not be reserved for the rich and famous”.

It is “less so for the practitioners, judges, and taxpayers, who ultimately bear the brunt of a greater volume of self-represented litigants”, they added, many of whose claims bear only a passing relationship to their prospects.

“For some, an AI-generated concerns notice may be the first step in vindicating their reputation, while for others, it will amount to little more than yelling into a pillow, only on a letterhead. Either way, the burden will fall on practitioners and courts to separate the two,” they said.

For complainants using AI, careful consideration must be given to case theory before sending a concerns notice, Parsons advised.

“When a concerns notice shows a disconnect between argument and fact, that can be a sign that AI has been used without this deeper consideration. For example, a concerns notice might elaborately assert serious harm to reputation, but on closer inspection, the supporting factual allegations might be negligible or misconstrued,” he said.

“Perhaps that’s simply because the user hasn’t given the AI tool enough context. This usually leads to expanding time and energy on both sides – the concerns notice recipient seeks further particulars, and the complainant is required to respond to keep the complaint alive. As costs accrue, it might become harder to resolve the dispute.”

Emerging issues and the need for human judgement

What concerns Snell about such developments is not the use of AI itself, but the “risk that people may have greater confidence in documents that look legally sophisticated, without appreciating the legal and strategic judgment required behind them”.

“The Defamation Act has consequences for those who fail to get their concerns notice right and those consequences cannot always easily be undone. Poorly framed concerns notices may also ultimately create more work for practitioners and the courts rather than resolving disputes earlier,” she said.

O’Connell said: “Those who rely on AI to bring defamation claims will more often than not struggle to be successful in court in the absence of experienced legal assistance. Even just using AI to send a defamation concerns notice (which is a mandatory first step in any defamation matter) is fraught, with the courts previously having thrown cases out on the basis of a defective concerns notices. In defamation, it is essential to get things right from the beginning.”

A concerns notice might appear straightforward, Snell continued, but identifying the imputations that genuinely arise, understanding the available defences and deciding strategically what outcome a client should pursue all require human judgement.

“AI can produce something that looks convincing, without necessarily getting those things right,” she said.

For lawyers, Snell remarks, the opportunity is to embrace AI for what it does well while recognising where human judgement remains essential.

“Used properly, AI can make legal services more efficient and accessible, but it should support, rather than substitute for, the judgment, strategy and responsibility lawyers bring to the administration of justice,” she said.

Looking ahead

Of course, O’Connell noted, it is understandable why people might turn to AI when legal costs are considerable. However, he said, “access to justice really only improves if quality keeps pace with accessibility. AI cannot offer the quality needed in defamation.”

For Parsons, while there’s a role for AI to save time and reduce costs in litigation more broadly, there’s a risk that, without timely legal advice that evaluates a case more holistically, case weaknesses are undetected until proceedings are well underway. “It’s difficult to predict, but for courts, this might lead to increasing interlocutory arguments and case management challenges,” he said.

Ultimately, Cutri and Bentley concluded, it’s too early to determine whether defamation proceedings need the safeguards now being introduced by the Fair Work Commission.

For now, the pair said, the immediate lesson for lawyers is simple: “AI will make it easier for aggrieved parties to articulate their concerns, but not necessarily to deal with them.”

“Clients will continue to put a premium on professional judgement, advice and strategy. The trick for practitioners will be proportionality. Not every AI-assisted notice should become an expensive dilemma for a client nor an all-consuming one for the lawyer.”

Want to see more stories from trusted news sources?
Make Lawyers Weekly a preferred news source on Google.
Click here to add Lawyers Weekly as a preferred news source.