P
Parachute
Parachute is a legal AI workspace for small and mid sized law firms from Veraty Tech Pty Ltd in Sydney, sold in Australia, New Zealand, the United Kingdom, the United States and Canada. It drafts documents from the firm's precedents, knowledge base and jurisdictions, and reviews contracts in several stages that identify obligations, flag conflicts with the firm's own policies and rate risks high, medium or low. Tabular review fills a grid across many documents, and research questions about Australian legislation come back grounded in the act and section with a link.
Three AI modes, Paralegal, Lawyer and Senior Associate, trade depth for credits. Around the AI sit matters and workstreams, a collaborative editor with track changes, email forwarded into threads, a meeting notetaker, client invitations and branded Word export. It connects to NetDocuments, Clio, Smokeball, InfoTrack, OneDrive and Slack, with a Word add in. Data is hosted in Australia, and Standard costs A$400 a month for the firm's instance and first seat, then A$120 a seat.
Parachute, head to head
Side by side on the same 15 capability axes and 12 legal signals: where the two part company, and what neither one publishes.
Alternatives to Parachute
Vendors in the same categories as Parachute whose graded profiles look most alike, with a line on what each product does.
Similar is not the same, so check that each one does the job you need. No vendor pays for placement.
- LLucioLucio is a legal AI workspace built by lawyers at Zenlegal Technology Private Limited in Bengaluru, which contracts outside India through Lucio Inc and has users in 13 jurisdictions.
- MMagnarMagnar is a legal AI agent for lawyers in Latin America, built by Magnar SpA in Santiago, Chile.
- LLiteraLitera is a Chicago legal technology company.
- VVesenceVesence is an AI platform for law firms and other professional services businesses that runs inside Microsoft Office.
- OOrdalieOrdalie is a French legal AI platform for lawyers in private practice, in house legal teams and notaries.
- GGistGist, formerly Bradwell, is a legal AI workspace and deployment service from Bradwell Tech, Inc., a Delaware company led by cofounders Shir Wisenberg and Tzuf Paz-Argaman.
Ordered by how closely the graded profiles match across the 15 capability axes, inside the same categories. See how each axis is graded in the methodology.
Capability grades
All 15 axes, graded from public sources on the date shown. Hover a grade to see what the letter means on that axis.
AI Centrality
How much of the product is actually AI. Whether the machine learning is the mechanism the buyer is paying for or a feature layered onto conventional software, and whether the vendor is specific about which is which.
Drafting and review are the product. A single prompt becomes a draft built from the firm's industry, jurisdictions, precedents and knowledge base, and sections can be refined with the model. Contract review runs a multistage analysis that identifies each party's obligations, flags conflicts with the firm's own policies and scores risks high, medium or low. Tabular review fills a grid of defined columns across many documents, and the notetaker joins Zoom, Teams or Google Meet calls to produce summaries, decisions and actions.
A model switcher and three modes route each task. Matters, workstreams, the editor and client management give the AI a workspace, but they are sold as one product with it.
Citation Accuracy and Hallucination Disclosure
Whether the vendor publishes measured accuracy on citations and assertions, grounds output to primary sources, and says plainly what its system does when it does not know. Legal has a documented public record of fabricated citations reaching filed briefs, so an untested claim of accuracy is not evidence.
Every answer shows its sources, so a claim traces back to the document, the legislation or the research behind it. Questions about Australian legislation are grounded in the act and section, with a link to the source. Review findings are compared with the firm's own knowledge base, so a clause that conflicts with a house policy is flagged with the policy it conflicts with. Parachute's own buying guide tells firms to expect clickable citations, live legislation and accuracy claims backed by a method, but no accuracy figure or test set of its own is published, and case law coverage is not described.
Autonomy and Oversight Model
What the system decides on its own, what a lawyer must approve, and whether the vendor documents where the review point sits. A tool that drafts under review and a tool that files without one are different products and different risks.
The terms treat AI content as information that may contain errors. It must not replace advice from a professional, and the privacy policy says the models assist and do not make final decisions without a chance for human review. In practice drafts open in an editor with track changes, comments and version history, and in line suggestions are accepted or rejected one by one. Firms can invite clients and act as the expert verification layer for their legal work.
The user picks Paralegal, Lawyer or Senior Associate mode, or lets Auto choose, and a daily briefing ranks the day's work each morning for the lawyer to act on. No threshold at which the product stops and hands back, and no route after a wrong output, is described.
Operational and Outcome Evidence
Named, dated evidence that the product works in production at real firms or legal departments. Case studies with figures and identified customers count. Unattributed testimonials and launch announcements do not.
Named customers give figures. Arisha Arif, senior legal counsel at Zed Law, says a 2.5 hour task took 20 minutes, and Simon Fletcher of Fletch Law says his lawyers save up to 10 hours a week. Founders at Nexa Legal, Zencheck AI and Locumate are also quoted. Parachute cites a 20 minute average contract review, down from two hours, and eight hours saved per lawyer each week. None of the figures comes with a date or a measurement method. InfoTrack became an integration partner in April 2026.
Privilege and Confidentiality Posture
How client confidences are handled: attorney client privilege and work product treatment, segregation of one client matter from another, whether client data trains any model, and what the vendor commits to in writing rather than in marketing.
Privilege is addressed directly. Submitting privileged or confidential material to the service does not waive privilege, under the privacy policy. Customer data is never used to train AI models, and the model providers, OpenAI, Anthropic and Google, are contractually barred from using, retaining or training on it beyond providing the service and are bound to confidentiality. Each organization's data is logically isolated, with role based access inside the firm.
The published terms add mutual confidentiality with notice before any compelled disclosure. Segregation between matters inside a firm is not described.
UPL and Professional Responsibility Posture
Whether the vendor is clear that it supplies a tool rather than legal advice, who its audience is, and how it addresses unauthorized practice of law, competence and supervision duties, and jurisdiction limits. ABA Formal Opinion 512 is the reference point. Where the advice line is not the duty a product raises, the axis is read through the nearest professional duty it does raise: judicial conduct rules and the reviewing duty for products sold only to courts, and the duty to bill for time actually spent for products that draft time entries.
The terms say plainly that the company is not a law firm and gives no legal advice, and that AI content is for information only and must not replace advice from a professional. Buyers include law firms, chambers, in house teams and businesses without lawyers. Where a business engages a law firm through the platform, the firm alone is responsible for the advice, and Parachute acts only as a limited agent for engagement and payment.
Service is limited to Australia, New Zealand, the UK, the United States and Canada, and sanctioned jurisdictions are excluded. Competence and supervision duties for lawyers are not discussed.
AI Governance and Bias Disclosure
Published governance over model behavior: who owns it inside the vendor, what is tested before release, and what is disclosed about disparate output across matter types, parties, or populations.
The privacy policy states that the models assist advisory work and do not make final, binding decisions without human review, and gives individuals the right to contest automated decisions. Secure development includes automated security scanning and peer review of every code change. No AI governance framework, release testing, accountable owner or analysis of uneven output is published.
AI Safety and Data Stewardship
Retention, deletion, access control, and what happens to prompts and documents after they are processed. Whether the vendor states its subprocessors and its incident practice, or leaves the buyer to assume.
Data is encrypted in transit and at rest, each organization's data is logically isolated, and production access follows least privilege. Automated backups and monitoring run continuously. An incident response plan sets escalation paths, and Parachute commits to notify affected customers promptly after a confirmed breach. The AI subprocessors are named as OpenAI, Anthropic and Google, each contractually barred from retaining or training on customer data.
After termination, customers have 30 days to export their data, after which it may be deleted, and trial or pilot data can be deleted at the end of the trial. No retention period applies to prompts during an active subscription.
AI Liability and Recourse
What the vendor stands behind contractually when its output is wrong. Indemnities, caps, carve outs, insurance, and whether any of it is published or only reachable through a negotiated agreement.
Under the terms of 21 September 2026, governed by New South Wales law, each party's liability is capped at the fees paid in the 12 months before the claim. Claims under the indemnities, breaches of confidentiality and prohibited uses carry a higher cap of ten times that amount, and gross negligence and intentional misconduct are not limited. Parachute defends customers against third party claims that the service infringes a patent, copyright, trademark or trade secret.
AI content is provided without warranty of accuracy, and neither party answers for losses from relying on it without independent checking. Trials and pilots are as is, with no indemnity.
Practice Systems Integration Depth
How deeply the product reaches into the systems legal work already lives in: document management such as iManage and NetDocuments, Word and Outlook, contract lifecycle management, matter management, e-billing, and court filing systems.
NetDocuments connects both ways: documents import into matters and work saves back to the firm's document system, a link released in June 2026. Clio and Smokeball connect as practice management systems, InfoTrack orders can be placed from the chat, and OneDrive files are reachable. A Word add in brings drafting and review into Word, with a refreshed task pane from April 2026. Emails forwarded to a firm inbox become AI threads with their attachments, and the notetaker joins Zoom, Teams and Google Meet. Slack is listed. What syncs with Clio and Smokeball, and in which direction, is not described.
Deployment Model and Data Residency
Where the software runs and where the data sits. Multi tenant cloud, single tenant, private deployment, on premises, and whether region of residence is a published option or an enterprise conversation.
Customer data is hosted in Australia. AI requests go to model providers in the United States by default, and customers that need every inference call to stay in Australia can have it processed on Australian infrastructure. Storage and model processing are therefore stated separately. Each organization is a logically isolated tenant on secure cloud infrastructure, and residency in other regions is on the roadmap. Which plans include Australian inference is not stated.
Security Certifications and Trust Center
Independent attestation a buyer can pull without a sales call: SOC 2, ISO 27001, penetration test summaries, a trust center with current reports and named scope rather than a badge image.
The information security management system is certified to ISO 27001:2022 by Compass Assurance Services, an accredited certification body, under certificate 8818-4353-01, which is available on request. Practices are aligned with SOC 2 controls without a SOC 2 report. Scope and certificate dates are not shown, and there is no trust center.
Model Supply Chain Disclosure
Which models sit underneath, whose they are, where they run, and whether the vendor commits to telling customers when that changes. A legal buyer inherits every dependency it cannot see.
OpenAI, Anthropic and Google are named as the large language model providers, with inference in the United States by default or in Australia on request. A model switcher released in February 2026 lets users pick a model, and routing assigns models to tasks; Enterprise customers get priority model access at peak times. The specific models are not named, and no notice before a change of model or provider is committed.
Commercial Transparency
Whether a buyer can learn what this costs without entering a sales process: published rates, the unit being charged, what sits behind an enterprise tier, and what implementation adds.
Standard costs A$400 a month for the firm's instance and first seat, then A$120 for each additional seat, with 20% off on an annual term billed monthly. Each plan includes a monthly allowance of requests. A Paralegal request counts as 1, a Lawyer request 3 and a Senior Associate request 15, and the allowance resets each period. Seats are pro rated when added, a 14 day trial is offered, and teams over 50 seats get volume pricing.
Enterprise adds SSO, dedicated support, custom onboarding and SLAs at a quoted price. The price is shown before any seat change is confirmed.
Firm and Practice Coverage
Who the product is actually built for. AmLaw, midlaw, small firm and solo, in house departments, government and courts, and which practice areas are supported rather than merely claimed.
Firms of one to 200 lawyers are the target, with versions for law firms, chambers, in house teams and businesses that buy legal work. Practice pages cover family, corporate, employment, litigation, migration, commercial and property law, and Australian legislation is the research source. Service is offered in Australia, New Zealand, the UK, the United States and Canada. Drafting covers documents such as privacy policies, contracts and terms of service from the firm's precedents. Practice areas outside its range are not stated.
6 public documents
The public pages on file for Parachute, with the recorded signals each one supports and the date it was last read. Open any of them and check the reading against the record.
-
goparachute.ai/platform2 signals
Primary Law Corpus Provenance, Court Disclosure Support
Read Oct 13, 2026
-
goparachute.ai/security2 signals
Client Data in Training, Ethical Walls and Matter Segregation
Read Oct 13, 2026
-
goparachute.ai/terms-and-conditions2 signals
Prompt and Output Retention, Third Party Request and Subpoena Notice
Read Oct 13, 2026
-
goparachute.ai/customers1 signal
Billing and Fee Posture
Read Oct 13, 2026
-
goparachute.ai/privacy-policy1 signal
Outside Counsel Guideline Readiness
Read Oct 13, 2026
-
Fabricated Citation Record
Read Oct 13, 2026
A$400 a month
- A firm pays A$400 a month to start, which includes one person.
- Each extra person costs A$120 a month.
- Paying for a year saves 20%.
- Each AI request uses credits, and deeper modes use more.
- Big firms get a custom price.
Standard is priced monthly in Australian dollars: A$400 covers the firm's instance and first seat, and each additional seat is A$120. Annual terms, billed monthly, save 20%. Each plan includes a monthly allowance of requests, counted as 1 for Paralegal, 3 for Lawyer and 15 for Senior Associate. Volume pricing applies over 50 seats, and Enterprise is quoted.
Implementation: Not stated for Standard. Enterprise includes custom onboarding at a quoted price.
Note: Monthly billing can be cancelled at any time. Seats added mid cycle are pro rated, and removed seats come off the next bill. Credits reset each billing period. A 14 day trial is offered, and trial data may be deleted at its end unless the customer subscribes or exports it.
Legal Signals
What each signal meansA signal records what public sources say on the date shown. It is not a grade and it is not a recommendation. Where a signal reads Not addressed, it means the index did not locate the material in public sources on that date, which is a statement about disclosure rather than about the product.
Client Data in Training
Can material a lawyer puts into this product be used to train a model?
A public policy or trust page states no training on customer content, with no matching term located in the published agreement.
Documents, knowledge base content and AI interactions are used only to answer for the customer's own organization and never to train AI models. The privacy policy adds that OpenAI, Anthropic and Google are contractually barred from retaining or training on customer data. The published terms themselves contain no training clause.
Prompt and Output Retention
How long does the product keep what a lawyer typed, and can that be set to zero?
Retention is acknowledged in public materials with no stated period.
After termination, customers have 30 days to export their data in a machine readable format, after which Parachute may delete it, and trial or pilot data may be deleted when the trial ends. Model providers may not retain customer data. No retention period applies to prompts and outputs during an active subscription, and no zero retention setting is described.
Ethical Walls and Matter Segregation
Does retrieval respect the firm’s ethical walls, or can the model read across them?
Segregation is asserted in public materials with no published detail on how it is enforced.
Each organization's data, knowledge base and AI interactions are separated from other organizations, and granular role based access limits what team members reach. Matters group threads, files and conversations. How access is restricted between matters inside a firm is not described.
Third Party Request and Subpoena Notice
If someone subpoenas the vendor for a firm’s data, does the firm hear about it first?
Terms commit to notice where lawfully permitted. No transparency report located.
The confidentiality clause of the terms allows disclosure required by law only with prior written notice, so the other party can object or seek a protective order. Separately, disclosure to courts and law enforcement, such as on receipt of a subpoena, is listed among the uses of personal information. No transparency report is published.
Primary Law Corpus Provenance
Where does the law in this product come from, and does the vendor have the right to use it?
Coverage is described by jurisdiction with no identification of the underlying corpus.
Research covers Australian legislation, with answers grounded in the act and section and linked to the source. The database behind it, case law coverage and the rights basis are not named, and drafting for other jurisdictions relies on the firm's own knowledge base.
Good Law Verification
Does the product tell you when the authority it just cited has been overruled?
No located public material addresses whether authority is checked for subsequent history.
Answers cite the legislation, document or research behind them, and Parachute's buying guide stresses live legislation over static training data. No check of whether a cited case or provision has since been overturned or amended is described.
Refusal and Uncertainty Behavior
What does the product do when the answer is not in the corpus?
No located public material addresses what the product does when it cannot ground an answer.
The terms warn that AI content may contain errors, omissions or outdated information, and every answer shows its sources. What the product does when it cannot ground an answer, whether it declines or flags uncertainty, is not described.
Fabricated Citation Record
Does a public court record exist addressing fabricated or hallucinated legal citations in output from this product?
No court order, opinion or disciplinary record addressing fabricated or hallucinated legal citations produced by this product has been located as of the date shown. This is a statement about the public record on that one subject, not a finding about the product, and this signal is not a litigation history.
The AI Hallucination Cases database maintained by Damien Charlotin, which records court decisions worldwide that address hallucinated AI content and the tool involved where known, has no entry naming Parachute or Veraty. Parachute attaches sources to every answer and tells users not to rely on AI content without professional review.
Bar Guidance Alignment
Has the vendor engaged in public with the ethics opinions its buyers are bound by?
No located public material engages with bar or ethics guidance.
Accuracy, sourcing and data security are covered in the terms and the buying guide, without reference to the professional conduct rules or law society guidance that bind lawyer customers in Australia, New Zealand, the UK, the United States or Canada. No ethics opinion is discussed, and no mapping of the product to such rules is published.
Billing and Fee Posture
Does the vendor address what happens to the bill when the work takes an hour instead of six?
Public materials claim time savings without addressing billing or disclosure, and the product sits inside a fee relationship between a lawyer and a client where those savings would change the bill.
Law firms buy Parachute, which cites eight hours saved per lawyer each week and contract reviews cut from two hours to 20 minutes. Firms can invite clients on retainer and bill through the platform's law firm engagements, and credits are counted per request. No guidance on billing or disclosing AI assisted work is published.
Outside Counsel Guideline Readiness
Can a firm get this vendor through a client’s AI clause without a bespoke negotiation?
A current subprocessor or model provider list is published.
OpenAI, Anthropic and Google are named as AI subprocessors, with inference in the United States by default or in Australia on request, and the other third parties that handle personal information are listed as well. Subprocessors owe confidentiality at least as protective as Parachute's own duty to the customer. The ISO 27001 certificate is available on request, and no client facing disclosure pack is published.
Court Disclosure Support
If a judge’s standing order requires an AI disclosure, can the product produce one?
Some elements of the record are available, short of a document level export.
Answers carry citations, and the editor keeps version history, track changes and threaded comments, with Word export that preserves them. The user can see which mode and model handled a request. No export ties a filing to the model, sources and reviewer, and no court disclosure template is published.