Product and intended use
Define the configuration, people, information and tasks being assessed.
Per-matter versus annual SaaS
Public claim only
This profile covers TrialView SaaS (login at trialviewlive.com) for court bundling, case preparation, AI analysis over the workspace and hearing presentation. Pricing is per matter or annual subscription. UK versus EU Azure region is a customer choice on the security page. No public MSA.
Ask: Confirm per-matter versus annual, UK versus EU Azure region, which AI features are on, and whether a signed MSA/DPA replaces the website privacy notice.
Azure UK or EU SaaS
Public claim only
Security page: data is hosted on Microsoft Azure UK/EU regions and never leaves the customer's regulatory zone without explicit consent. AI processing is stated to occur exclusively in Azure data centres in the UK or the EU. No Australian or US region is offered on that page. On-premises is not described.
Ask: Which Azure region (UK or EU) will store our workspace, backups and AI inference, and is an Australia region available?
Bundles, AI Q&A, hearings
Public claim only
Homepage and AI page: hyperlinked court-compliant bundles (pagination, indexing, cross-references); timelines and pattern detection; natural-language questions over the entire case portfolio with links to sources; comparison of witness statements for contradictions; evidence presentation, video conferencing and real-time transcription. Workspace permissions include confidentiality rings.
Ask: Which of bundling, AI chat, inconsistency analysis, transcription and hearing presentation will be enabled, and which remain off?
Irish company; UK/EU hosting; no AU region
Public claim only
TrialView Limited is Irish-registered (Dublin). Privacy policy is GDPR and Irish Data Protection Acts 2018. Security page limits hosting and AI processing to Azure UK or EU. An Australian firm would send matter data to the UK or EU unless a later contract says otherwise. No Australian primary-law research feature was found.
Ask: Will our matters be stored and inferred only in the UK or only in the EU, and what APP 8 wording do you provide for an Australian client?
Data, controls and requirements
Compare the actual configuration with firm, client and jurisdiction requirements, supported by evidence.
Confidential case workspace; UK/EU transfer
Not confirmed
TrialView holds litigation/arbitration documents. Privacy: client is controller, TrialView is processor (Art. 28 GDPR) unless otherwise agreed. Australian clients would be sending matter data to Azure UK or EU. Confidentiality rings are a permission feature. No per-client AI exclusion was documented.
Ask: Can we disable AI analysis per matter, and what UK/EU region description can we give clients?
Azure OpenAI and Anthropic on Azure
Public claim only
Security page: TrialView uses Azure OpenAI and the Anthropic Service; currently the latest models from OpenAI and Anthropic, with a model-agnostic architecture for upgrades. Azure OpenAI is stated not to interact with OpenAI-operated ChatGPT. All AI processing in Azure UK or EU. Specific model names (for example GPT-4o) were not listed.
Ask: Which Azure OpenAI and Anthropic model versions and regions process our workspace, and will you notify us before changing them?
List available on request
Not confirmed
Privacy s.6.4: sub-processors (including secure hosting or infrastructure) are under Article 28 DPAs; a list may be made available on request. Azure (Microsoft) and Anthropic are named on the security page. No public dated list, countries or notice/objection period was published.
Ask: Provide the current subprocessor list (Azure, Anthropic, transcription, video, support) with countries and 30-day notice/objection.
No training except written agreement or anonymised data
Public claim only
Security page: prompts, completions and embeddings are isolated; not used to train TrialView's model; never used to improve OpenAI, Microsoft or third-party models; never used to train public AI models. Privacy s.6.2: Client Data is not used to train, fine-tune or improve AI models except where explicitly agreed in writing with the client or where data has been irreversibly anonymised.
Ask: Confirm the signed DPA has no written-agreement or anonymisation carve-out unless we opt in, covering Azure OpenAI and Anthropic.
Anonymised-improvement carve-out
Public claim only
The only published improvement path is privacy s.6.2's exception for irreversibly anonymised data or written client agreement. In-product feedback buttons were not described. Website analytics are a legitimate-interest purpose for website data, not Client Data.
Ask: Is there any in-product feedback or evaluation use of our documents or prompts, and is anonymised improvement off unless we agree in writing?
As long as necessary; no day count
Public claim only
Privacy s.9: personal data retained only as necessary for contract, law or claims, then securely deleted or anonymised. No numbered schedule for live workspaces, prompts, embeddings or backups.
Ask: Provide a numbered retention schedule for documents, prompts, outputs, embeddings, logs and backups while a matter is live.
Delete or return at client's election
Public claim only
Privacy s.6.5: on termination or expiry, Client Data will be deleted or returned at the client's election, subject to legal or regulatory retention. No day count or certificate is published. No public MSA restates this.
Ask: On matter close or contract end, what is deleted from Azure, backups and Azure OpenAI/Anthropic, within how many days, with what certificate?
Authorised personnel for support and maintenance
Public claim only
Privacy s.6.3: access limited to authorised personnel under contractual confidentiality who require access solely for service delivery, maintenance or support. Compelled-disclosure notice to the customer is not stated.
Ask: Is support access per-ticket with our approval, logged to us, and will you notify us of legal process seeking matter content?
Customisable permissions and confidentiality rings
Public claim only
Security page: workspace permissions can be customised for levels of access and confidentiality rings. Audit logging is listed among security measures. SAML/SSO, MFA enforcement and SCIM were not specified.
Ask: Do you support SAML/OIDC SSO-only, MFA enforcement, SCIM and confidentiality rings that the AI cannot cross?
TLS 1.2+ and AES-256 claimed across all layers
Public claim only
Security page: all data encrypted in transit using TLS 1.2+ and at rest using AES-256, with no exceptions across any data layer. Privacy also lists encryption. Customer-managed keys were not mentioned. Claims are not in a downloaded SOC 2 report.
Ask: Confirm TLS 1.2+ and AES-256 for Azure storage, backups and Azure OpenAI/Anthropic paths, and whether CMK is available.
No named SOC 2 or ISO certificate on the security page
Not confirmed
The security page describes Azure residency, encryption and AI isolation but does not name SOC 2, ISO 27001 or a public report. A whitepaper is offered on request via info@trialview.com.
Ask: Provide current SOC 2 Type II or ISO 27001 reports covering TrialView and Azure OpenAI/Anthropic processing in our region.
No public MSA; processor terms in privacy notice
Not confirmed
No public terms of service, liability cap or DPA URL were found. Privacy policy (27 January 2026) states TrialView is an Article 28 processor of Client Data, Irish controller for its own data, Irish DPC as complaints body. Updates to the privacy policy may be posted on the website.
Ask: Will our MSA include an Article 28/APP DPA, output ownership, a meaningful cap for confidentiality/data breach, and Australian or Irish law/forum?
UK/EU Azure; SCCs if outside EEA
Public claim only
Security: data never leaves the UK/EU regulatory zone without explicit consent; AI processing only in Azure UK or EU. Privacy s.8: transfers outside the EEA use adequacy or Commission SCCs. Azure OpenAI is described as not sending data to OpenAI-operated ChatGPT. An Australian customer is outside the EEA, so APP 8 and SCCs/adequacy need a written map.
Ask: Name every country of storage and inference for our tenancy, including Anthropic-on-Azure, in writing.
No public notification period
Not confirmed
Privacy lists security measures but does not state a numbered customer-notification period or NDB clause. Security and data-controller mailboxes are published.
Ask: Within how many hours of discovering a suspected incident affecting our Client Data will you notify us, and will you support an NDB assessment?
Model-agnostic upgrades; privacy may be updated
Public claim only
Security: model-agnostic architecture to allow secure upgrades as improved models become available. About page: more features and upgrades over the next year. Privacy may be updated from time to time on the website. No public change-notice period for subprocessors.
Ask: Will our MSA freeze terms during the term, and notify us before changing Azure OpenAI/Anthropic models or UK/EU regions?
Delete or return at election; format unnamed
Public claim only
Privacy s.6.5: delete or return Client Data at the client's election on termination, subject to legal retention. No bulk-export format, bundle/PDF package or certificate is specified. Per-matter packages imply a natural matter-end; annual subscriptions were not described for exit.
Ask: How do we export bundles, transcripts, AI chat logs and audit trails on exit, in what format, and when are remaining Azure copies deleted?
No workplace connectors documented
Not confirmed
No Clio, iManage, Relativity or email connectors were listed on the pages reviewed. The workspace appears to be fed by customer upload. Hearing services include video conferencing, which is a processing path.
Ask: What systems can ingest into a workspace, and what does the hearing/video stack send outbound?
Analysis limited to the case workspace
Public claim only
AI FAQ: analysis across the entire case portfolio in the workspace; chat searches the document portfolio and returns linked sections. Open-web search as a TrialView feature was not found. Security: analysis remains strictly within the private case workspace.
Ask: Confirm TrialView AI does not call the open web on our matters.
Q&A and analysis inside the workspace
Public claim only
AI features described are search, Q&A, inconsistency detection and bundle automation inside the workspace. No MCP or agentic write-out to email/court e-filing was documented. Hearing presentation can display evidence live.
Ask: Which actions (bundle generation, export, hearing display) can run without a named approver, and how do we freeze AI chat immediately?
Links to source; no contractual review duty
Public claim only
Chat is described as directing the user to the answer plus other relevant linked sections for review. No public terms clause requires a qualified legal professional to approve outputs. The product is marketed to practising litigators who remain responsible for court use.
Ask: Can we require named sign-off before a bundle or AI-derived chronology is exported, with a log of who verified which items?