Built By Barristers, Australia’s New AI Tool Takes On Legal Research Giants

Matthewkalyk lawfuel

Australia’s increasingly crowded legal AI market has another challenger, with Sydney-based BarristerAI has launching an AI legal research platform built specifically for barristers and smaller practices, charging $30 a month plus usage fees rather than the substantial enterprise subscriptions associated with many eokstablished legal-tech platforms.

Founded by barristers Matthew and Frank Kalyk, (Matthew pictured above) the company claims its system outperformed Anthropic’s powerful Claude Fable 5 model in its own Australian legal-research benchmarking, while surfacing substantially more relevant authorities.

The timing is useful. The Federal Court of Australia’s new Generative AI Practice Note expressly recognises AI’s potential to cut litigation costs and improve access to justice, while demanding that lawyers verify authorities, protect confidential information and remain responsible for everything AI produces.

Frank kalyk

Matthew Kalyk practices as a barrister in Sydney which he has done so for over a decade and was formerly with the Director of Public Prosecutions. His co-founder Frank Kalyk (above) is a retired barrister with 35 years’ experience and currently grows organic merlot in Bordeaux, along with helping oversee the new legalAI tool.

The firm’s media release on the tool is below –


Banner Image
BarristerAI launches in Australia, using AI to expand access to justice
facts:
BarristerAI, a legal research platform built specifically for barristers, has launched in Australia, offering AI-powered legal research at significantly lower cost than established competitors such as LexisNexis, Thomson Reuters, and Harvey.
The platform has benchmarked its performance against Fable, described as the world’s most powerful AI model, claiming 10% better legal reasoning and approximately 20 times more legal authorities surfaced, with all outputs including pinpoint citations linked directly to their sources.
Confidentiality is a core design principle, with research queries able to be anonymised before reaching any AI provider, no use of queries for external model training, and a zero-data-retention Private Memo option available for privileged matters.

The platform includes features such as personalised case feeds filtered by practice area, case citation history, and Smart Legislation enabling users to interrogate statutory provisions and retrieve legislation as it stood at any given historical date.
Pricing is structured to be accessible to sole practitioners and small practices, at $30 per month plus pay-as-you-go costs of $1 for quick searches and $25 for full research memos, with a no-commitment trial available for new users.

BarristerAI launches in Australia, using AI to expand access to justiceBuilt by barristers.
Benchmarked performance above the world’s most powerful AI model. Affordable to any barrister or sole practitioner.


SYDNEY,  17 August 2026  BarristerAI, an AI-powered legal research platform built specifically for barristers is today launching their new legal AI research product at a fraction of the cost of the subscriptions that have long dominated the market.The law belongs to everyone; so should the best AI research tools.
Yet the available products either come at a price of multi-thousand dollar annual subscriptions aimed at large firms ( think LexisNexis, Thomson Reuters, Harvey, Legora) or off the shelf AI chatbots such as Claude and ChatGPT which are not built or verified for legal work. 
BarristerAI is a third option: purpose-built legal research by barristers, providing the accuracy safeguards necessary for a professional tool but user friendly and priced like a consumer subscription.
We went head to head against the best publicly available model in the world, Anthropic’s Fable.  
The result?  Better legal reasoning and more relevant authorities surfaced.  All with pinpoint references.  A product lawyers can actually use. “We want to provide the best legal research platform on the market and have it accessible by every lawyer in Australia,” said founder Matthew Kalyk, a barrister practicing from 12 Wentworth Selborne Chambers in Sydney. “We didn’t set out to build a cheaper research tool.
We set out to build the best one. It just happens to be cheap because of its architecture, engineering and functionality.”
Matthew Kalyk says the idea began with his own experience with AI: needing first-rate legal research and finding the subscriptions that provided it were just not meeting the level of quality reliability that he needed as a barrister.
Frank Kalyk, co-founder of Barrister AI practised law for more than 35 years; as a commercial litigation solicitor, then at the New South Wales Bar.Every one of those years ran on research: long hours and late nights just to find and read the authorities, largely to set aside the irrelevant, to distill the critical and to pinpoint the relevant references.
“It is critical for barristers to have accurate research of the law. It is also a basic right of all members of our society to have access to justice, which requires access to the law that is affordable. For too long now, that access has been out of the reach of the normal person, because of the prohibitive costs. AI gives us a chance to change this,” Frank said.
“But if tools like this are only available to big law, then the problem of prohibitive costs will only be perpetuated and  we cannot have fair access to justice. This is an opportunity to provide access to justice to all, not only the well funded .”
Why barristers need something different
The company argues that generic AI tools, however capable, do not approach the law in the way barristers must actually work. “At the bar, it’s the reasoning by reference to the authorities that matters,” Frank said.
“An answer with no reasoning behind it is useless to a barrister. Most general AI tools hallucinate, or are priced for large solicitors’ firms, not advocates.”
This means that small practitioners or other interested parties do not have access to those resources. BarristerAI, he says, redresses this. It is built specifically for practitioners who need precision and sources they can check line by line, without paying large subscription fees.
Benchmarked against the world’s leading AI model
BarristerAI is platform-agnostic, meaning the company continuously tests leading AI models from different providers and runs its research pipeline on whichever performs best, rather than building around a single one. 
On its benchmark of difficult Australian legal research, BarristerAI says its research product outperforms Fable, currently regarded as the most powerful AI model in the world by 10% on legal reasoning and surfaced around 20 times more authorities. It produces a legal memo with pinpoint references. 
“That margin is not even the best about the BarristerAI platform,” Frank said. “It’s the product of thoughtful design.”
Built for verification, not just generation
Every proposition in a BarristerAI research output is grounded in and linked to its source, with pinpoint citations that can be checked with a single click, a design that BarristerAI  says enables verification in compliance with the Practice Notes of Courts around Australia on the use of generative AI in litigation. 
Confidentiality is treated by BarristerAI as an essesstial obligation rather than a feature add-on: research queries can be anonymised before reaching any AI provider, are never used to train external models, and a zero-data-retention “Private Memo” option is available for privileged matters.
The platform also includes a personalised feed of the latest cases and legal news filtered to each barrister’s practice areas; consistent headnotes and “QuickTake” summaries generated across its entire case database; “Smart Cases,” which show a case’s full citation history at a glance   whether it has been followed, distinguished or doubted, and why; and “Smart Legislation,” which lets a user interrogate any provision down to the section, sub-section and definition to see how courts have construed it, trace its amendment history, and retrieve a provision as it stood at any given date   the version that governed a client’s dispute, not the version in force today   with quick summaries and deeper analysis available at a click.
“The test I set was simple,” Frank said.
“Could I have usefully relied on this at the Bar? Would I have put it before a judge? Until the answer was yes, we weren’t satisfied.”
Pricing built for solo and small practices
BarristerAI does not believe in up-front costs that lock in customers irrespective of the value being provided.  The product costs $30 a month per barrister and then is pay as you go: Quick Searches at $1 and full Research Memos at $25, with a metered option for experienced users who can bring memo costs down further.
New users get a trial with no platform fee and no commitment, letting them test the platform on their own matters, in their own time. The model is designed for barristers, Frank says, but it means a sole practitioner in any regional town in Australia now has the same research power as a national firm. “That,” he said, “is access to justice.”
An open challenge
“This is a space where lawyers need to try the product to see how good it is,” Matthew said. Even compare it to other products on the market.  BarristerAi invites barristers to submit their most difficult legal research questions directly to BarristerAI, so they can compare its output against other tools and judge the results for themselves.BarristerAI is available now at barristerai.com.au.
About BarristerAIBarristerAI is an Australian legal technology company providing AI-powered legal research and related tools for barristers and small legal practices. Founded by Frank Kalyk, a litigation solicitor and barrister of more than 35 years, the platform is built to make frontier-grade legal research accessible and affordable to sole practitioners and small firms, not just large enterprises. BarristerAI is independent and Australian-owned.

Leave a Comment

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Scroll to Top