Early access Built country by country
Legal AI that shows its sources.
Polylex AI is building a legal assistant for each country, drawn only from official and licensed sources. Every source it returns links to the official text, and your client files stay private.
- Official and licensed sources only
- Every source linked to the official text, with its status and date
- Your files are never used to train AI models
Polylex Research
Australia · Commonwealth
Question: What does APP 11 require of an entity that holds personal information?
Answer: It must take such steps as are reasonable in the circumstances to protect the information from misuse, interference and loss, and from unauthorised access, modification or disclosure.Source 1 Those steps include technical and organisational measures.Source 1
Privacy Act 1988 (Cth), Schedule 1, APP 11
Open the official text11.1 If an APP entity holds personal information, the entity must take such steps as are reasonable in the circumstances to protect the information: (a) from misuse, interference and loss; and (b) from unauthorised access, modification or disclosure.
Sourced from the Federal Register of Legislation at 4 October 2026. For the latest information on Australian Government law please go to https://www.legislation.gov.au. Licensed under CC BY 4.0.
What we’re building
One platform. A legal product for each country.
-
Research
Legislation, regulations and case law, each source linked to its official text.
-
Drafting
Documents in the formats local courts expect.
-
Case files
Upload files for OCR, fact extraction and search by file and page.
-
Calculators
Local calculations such as severance, interest and indexation.
-
Where you work
In AI assistants that support MCP connectors, and in Microsoft Word.
ISources
Official and licensed sources only.
We take the law from the bodies that publish it, such as courts, legislatures and official gazettes, and from publishers licensed to distribute it.
- Cleared before collection
- Every source is recorded in our source registry before anything is collected: who publishes it, under what licence, what its terms of use and robots rules allow, and the evidence, dated.
- The right way in
- An official API or open-data release first; then a licensed feed; and only then careful collection of public pages, where the terms allow it.
- What we never use
- Competitors’ databases, even through an open API, and anything a source’s terms forbid. Not even for testing.
-
- Source
- Official legislation portal
- Access
- Official API
- Licence
- Open government licence
- Status
- Approved
-
- Source
- Supreme Court judgments
- Access
- Bulk download
- Licence
- Terms under review
- Status
- To verify
-
- Source
- Official judgment publisher
- Access
- Licensed feed
- Licence
- Commercial licence
- Status
- Licence needed
-
- Source
- Commercial legal database
- Access
- None
- Licence
- Competitor
- Status
- Prohibited
Status lifecycle: CandidateTo verifyApproved, licence needed or prohibited
IIVerification
From the official text to your answer.
A citation is only as good as the text behind it. Each step keeps the link back to the official source.
-
Collected
Taken from an approved source. The original file is archived unaltered.
-
Read
Text extracted, with OCR for scanned documents, and structured by jurisdiction, issuing body, date and official identifiers.
-
Versioned
Legislation kept as in force on each date. Where no official consolidation exists, ours is labelled unofficial.
-
Returned with its source
Every source Polylex returns carries a link to the official text, its status and the date retrieved. A source we can’t match is flagged.
Definitions
On this website:
- “official source”
- means a court, legislature, gazette or other public body that publishes a legal text, or a publisher it licenses;
- “licensed source”
- means a source we use under a written licence;
- “verified”
- means matched to the text at its official source, with the date retrieved;
- “in force”
- means the status the official source gives a text on the date shown;
- “unofficial consolidation”
- means a point-in-time text we compile where no official one exists, always labelled as such.
IIIPrivacy
Your files stay yours.
Lawyers keep their clients’ confidences. Polylex is designed around that duty.
-
No training on your data
We don’t use your questions, documents or case files to train AI models.
-
Not for sale
We don’t sell your data or use it for advertising.
-
Encrypted
Files will be encrypted in transit and at rest.
-
Hosted by region
EU customers’ files will be hosted in the EU. For other countries, we’ll publish where data is hosted before launch.
-
Kept separate
Each firm’s files will be kept apart, with access limited to the people the firm authorises.
-
Yours to delete
Delete a file, or your whole account, at any time.
-
People named in court decisions
We keep the anonymisation courts apply, never re-identify parties or link them across sources, and review and act on every takedown request. How to request a takedown
Your AI assistant keeps its own chat history, under settings you control. What Polylex receives
IVCountry by country
A product for each country, not a translation.
Each country gets its own product, in its own language, built around its own legal tradition: civil, common or mixed.
Map the legal system
The hierarchy of norms, the courts, which decisions bind, and how lawyers cite.
Clear the sources
Licence, terms and evidence, recorded for every source.
Collect and check
Coverage measured source by source.
Build local tools
Drafting in local court formats, and local calculators.
Pilot with local lawyers
Practising lawyers use it first and tell us what is missing.
Launch with a coverage statement
What is included, from which official source, from which date.
Status: no country is live yet. We are preparing the first jurisdictions now.
Who we are
Independent by design.
Polylex AI is an independent legal‑technology company. We are self‑funded, so we answer to the lawyers who use our products, not to investors.
Each country’s product is built with practising lawyers from that country: the people who know which sources matter and how their courts expect to see them cited.
Is Polylex available in my country?
Not yet. No country is live. We are preparing our first jurisdictions and will open them one at a time. Request early access and tell us where you practise; we will write when your jurisdiction opens.
Where does the legal content come from?
Only from official sources, such as courts, legislatures, official gazettes and other public bodies, and from publishers licensed to distribute their texts. Each source is cleared for its licence, terms of use and robots rules before anything is collected.
Can the assistant still get a citation wrong?
Yes, it can. Polylex works inside your AI assistant, and the assistant writes the final answer. What we control is what Polylex returns: every source comes with a link to the official text, its status and the date retrieved, and we flag any source we can’t match.
Check the sources before you rely on them. You remain responsible for what you file.
Do you train AI models on my files?
No. We don’t use your questions, documents or case files to train AI models, and we don’t sell your data.
What about the AI assistant I use?
Your assistant keeps its own chat history, under its own settings, which you control. Polylex receives only the requests your assistant sends to it and the files you choose to upload to Polylex.
For client work, we recommend an assistant plan that does not train on your conversations.
Where will my data be stored?
Files will be encrypted in transit and at rest. EU customers’ files will be hosted in the EU. For every other country, we will publish where data is hosted before that country launches.
Do you use other legal databases?
Never a competitor’s database, even where an open API would allow it. We use official sources and licensed feeds only.
What about personal data in court decisions?
We keep the anonymisation courts apply, and we never re-identify parties or link them across sources. Anyone named in a decision can ask us to remove or anonymise it. We review every request and act on it, following the official source’s own anonymisation and removals. How to request a takedown.
Is this legal advice?
No. Polylex is a research and drafting tool for lawyers. It is not a law firm and does not give legal advice.
Early access
Request early access.
No country is live yet. Tell us where you practise and your practice area, and we’ll write once, when your jurisdiction opens. No newsletter.
Email hello@polylexai.comPlease include your name, country, practice area and firm size.
Publish the law?
We would rather license your texts than work around you. Licensed sources are credited, linked back to your official record and used only as the licence allows.
Talk to us about licensing