Scope & Capabilities
What AI-assisted features actually do — and what counts as an approved capability versus an illustrative example.
Jump to section →LEGAL / AI TERMS
AI Terms
Terms governing how Talvrin's AI-assisted features may be used — and what to verify before relying on anything they generate. AI helps you navigate evidence; it does not become the evidence.
Version 1.0
Effective and last-updated dates publish once Legal confirms governance records.
Version history — not yet available
AI AT A GLANCE — CONVENIENCE SUMMARY ONLY
Six clause families. The sections below remain controlling.
AI assists research; generated output is not authoritative evidence; verify sources before relying on a summary; human judgment remains responsible; and nothing here is investment advice or trade execution. This summary is a convenience orientation — if approved policy is broader than what is shown here, the full sections below govern.
Evidence Boundary
AI output is assistance, not authoritative evidence — the underlying source stays separately inspectable.
Jump to section →Inputs & Outputs
What you may submit as a prompt, and how you may reuse or represent what AI returns.
Jump to section →Governance & Changes
Prohibited use, IP, enforcement and version changes stay sourced from approved registries and Legal review.
Jump to section →Your Responsibility
Verify sources, apply human judgment, and treat non-advice and non-trade-execution boundaries as fixed.
Jump to section →Privacy & Data Rights
Prompt, output, retention, training and market-data-rights facts stay governed by approved policy, not marketing copy.
Jump to section →
SCOPE & CAPABILITIES
1. What these AI Terms cover
These AI Terms govern your use of Talvrin’s AI-assisted features and apply together with Talvrin’s Terms of Service, Privacy Notice, Acceptable Use policy and Data Processing Addendum. They do not replace those documents — where a specific AI behavior is also addressed elsewhere, the more specific, currently approved term controls.
Related:Terms of Service →Acceptable Use →
SCOPE & CAPABILITIESNO APPROVED RULE PUBLISHED YET
2. What Talvrin AI-assisted features are
AI-assisted features can help with discovery, organization, comparison, summarization, identifying changes and explaining relationships within the evidence Talvrin presents. These are illustrative capability categories, not a commitment that every feature is available to every account or in every market.
Talvrin has not supplied an approved AI Capability Registry describing exact feature availability by account tier or region. This page does not claim a feature is live unless Talvrin has actually released it; your product experience is the current source of truth for what is available to you.
ILLUSTRATIVE EXAMPLES — NON-EXHAUSTIVE
- Searching and organizing evidence across sources, filings and market data.
- Summarizing or comparing related evidence at your request.
- Identifying and explaining changes between versions of the same evidence.
Related:AI Assistance →Platform Overview →
EVIDENCE BOUNDARY
3. AI is assistance, not authoritative evidence
AI should help people navigate evidence — it does not replace it. Generated output does not become authoritative simply because it reads as fluent or persuasive, and it must remain visually and semantically distinguishable from the underlying source evidence.
This is Talvrin’s core AI doctrine. Every other section of these AI Terms operates within this boundary.
EVIDENCE BOUNDARY
4. Evidence, provenance and source inspection
Where AI-assisted output relates to specific evidence, the underlying source remains separately inspectable. Review the original evidence a summary, comparison or explanation is drawn from, rather than treating the AI-generated text itself as the record.
Related:Evidence →Evidence Standards →
YOUR RESPONSIBILITY
5. Human judgment and user responsibility
Human judgment remains responsible for investment, professional and organizational decisions. Before relying on an AI-assisted interpretation for a material decision, inspect the underlying evidence rather than the generated summary alone.
ILLUSTRATIVE EXAMPLES — NON-EXHAUSTIVE
- Inspecting underlying evidence before relying on an AI-assisted interpretation for a material decision.
- Treating an AI-generated explanation as a starting point for research, not a finished conclusion.
- Applying your own professional or organizational review process before acting on AI-assisted output.
YOUR RESPONSIBILITY
6. No investment advice or trade execution
Talvrin does not position itself as a brokerage, trade-execution service, stock-tipping service or recommendation engine — including through its AI-assisted features. Nothing generated by Talvrin’s AI constitutes investment, legal, tax or other professional advice, or a recommendation to buy, sell or hold any financial instrument. Talvrin does not promise to predict markets or eliminate uncertainty.
This boundary applies to AI-assisted output the same way it applies to the rest of Talvrin’s research and intelligence platform.
YOUR RESPONSIBILITYNO APPROVED RULE PUBLISHED YET
7. Accuracy, uncertainty and completeness
AI-assisted output may be incomplete, outdated or inaccurate. Talvrin does not represent that generated summaries, comparisons or explanations are complete or error-free, and does not assign or display a numeric confidence or accuracy score for AI-assisted output.
Talvrin has not approved a methodology for quantifying AI output confidence or accuracy. Until one exists, no percentage, score or similar certainty indicator is published for AI-assisted features.
INPUTS & OUTPUTS
8. Prompt and input responsibilities
Only submit prompts or inputs you are authorized to provide, and only in ways consistent with Talvrin’s Privacy Notice, Acceptable Use policy and, where applicable, Data & Market Information Terms. Do not submit content you do not have the rights or permission to share.
ILLUSTRATIVE EXAMPLES — NON-EXHAUSTIVE
- Submitting another person’s confidential or personal information without a lawful basis to do so.
- Pasting licensed market data or third-party content into a prompt beyond what your entitlement permits.
- Using prompts to attempt to extract another user’s account data or Talvrin’s internal systems.
Related:Acceptable Use →
INPUTS & OUTPUTS
9. Output use and downstream provenance
Review AI-assisted output before reusing it, and do not misrepresent generated text as an official source, issuer statement, regulator communication or primary evidence. Where an output conflicts with the underlying evidence, the evidence record controls, subject to its own provenance and version state.
ILLUSTRATIVE EXAMPLES — NON-EXHAUSTIVE
- Citing the original evidence rather than an AI-generated summary when provenance matters.
- Disclosing that a shared summary or comparison was AI-assisted when repurposing it outside Talvrin.
PRIVACY & DATA RIGHTSNO APPROVED RULE PUBLISHED YET
10. Privacy, data handling and retention
Privacy, data handling and retention for AI-assisted features are governed by Talvrin’s Privacy Notice and Data Processing Addendum, not by this page. Where this page and those documents overlap, the approved Privacy Notice and Data Processing Addendum control the operative data-handling facts.
Talvrin has not published a dedicated AI data-flow registry describing prompt- and output-specific retention periods. Until one is approved, the general retention and data-handling commitments in the Privacy Notice and Data Processing Addendum apply.
Related:Privacy →Data Processing Addendum →
PRIVACY & DATA RIGHTSNO APPROVED RULE PUBLISHED YET
11. Model providers and subprocessors
Model providers and subprocessors used to deliver AI-assisted features are disclosed here only once Talvrin approves a public model/provider record.
Talvrin has not supplied an approved AI model/provider registry for public disclosure. This page does not name a model provider, model family or version until Legal and AI Governance approve that fact for publication.
PRIVACY & DATA RIGHTSNO APPROVED RULE PUBLISHED YET
12. Training and model improvement
Whether prompts, inputs or outputs are used to train or improve Talvrin’s AI-assisted features is stated here only as an approved factual position.
Talvrin has not published an approved position on training or model-improvement use of prompts or outputs. No claim — positive or negative — is made on this page until Legal and AI Governance approve one; the Privacy Notice and Data Processing Addendum remain the controlling data-use documents in the meantime.
PRIVACY & DATA RIGHTS
13. Confidential, personal and restricted information
Do not submit personal data belonging to others, confidential information you are not authorized to share, or restricted content outside your entitlement as a prompt or input to an AI-assisted feature. Privacy and security rules that apply elsewhere on Talvrin apply equally to AI inputs.
ILLUSTRATIVE EXAMPLES — NON-EXHAUSTIVE
- Entering a third party’s personal or health information into a prompt without a lawful basis.
- Submitting confidential workspace or client information beyond your authorization.
Related:Privacy →Security →
PRIVACY & DATA RIGHTSNO APPROVED RULE PUBLISHED YET
14. Market data, licensed content and source rights
Licensed market data and other rights-sensitive source content remain subject to their original licensing terms when used as an AI input or reflected in AI-assisted output. Using a source as an AI prompt does not expand your rights to copy, extract or redistribute it.
Service- and dataset-specific AI rules for licensed market data are governed by Talvrin’s Data & Market Information Terms, which remain blocked pending approval of items 01–06 and this AI Terms item. Until published, your existing data license or entitlement governs.
Related:Data Rights →Acceptable Use →
GOVERNANCE & CHANGESNO APPROVED RULE PUBLISHED YET
15. Prohibited AI use
Do not use Talvrin’s AI-assisted features to misrepresent authorship or evidence provenance, fabricate or misattribute source material, or present AI-generated text as independently verified fact.
A complete AI-specific risk and misuse taxonomy has not yet been approved beyond the rules stated here and in Talvrin’s Acceptable Use policy. Additional prohibited-use categories publish only once Legal and AI Governance approve them.
ILLUSTRATIVE EXAMPLES — NON-EXHAUSTIVE
- Presenting AI-assisted output as independently verified evidence without disclosure.
- Using AI features to fabricate or misattribute source material.
- Relying on AI-generated text as a substitute for reviewing the underlying evidence.
Related:Acceptable Use →
GOVERNANCE & CHANGESNO APPROVED RULE PUBLISHED YET
16. Automated decisions and high-risk use
Rules for automated-decision or high-risk AI use apply only for categories Talvrin’s Legal and AI Governance teams have specifically approved.
No automated-decision or high-risk AI use has been approved for publication on this page. Talvrin’s AI-assisted features are positioned as research assistance, not as automated decision-making authority.
GOVERNANCE & CHANGESNO APPROVED RULE PUBLISHED YET
17. Intellectual property and content rights
Rights in prompts, AI-assisted outputs and related feedback are addressed here only once Legal approves a jurisdiction- and deployment-specific position.
No approved prompt/output ownership, licensing or non-infringement claim is published here. Do not infer ownership or reuse rights from the availability of AI-generated output.
GOVERNANCE & CHANGES
18. Availability, beta and feature changes
AI-assisted features, including any non-production or beta functionality, may change, be limited or be withdrawn consistent with Talvrin’s approved terms. A feature described as available is only as available as Talvrin’s current product release state supports.
Beta or limited-availability AI features are labeled as such in the product where applicable; this page does not promise permanence for any specific AI capability.
GOVERNANCE & CHANGESNO APPROVED RULE PUBLISHED YET
19. Monitoring and safety controls
Any monitoring or safety controls applied to AI-assisted features are described here only to the extent Talvrin has approved public disclosure of them.
Talvrin has not approved specific AI monitoring or safety-control disclosure details for publication on this page. This section does not claim that a particular monitoring mechanism is present.
GOVERNANCE & CHANGES
20. Enforcement and suspension
Talvrin may take account-level action in response to a violation of these AI Terms or the Acceptable Use policy, consistent with the Terms of Service.
The specific enforcement actions, triggering conditions and any related review process are governed by the applicable Terms of Service and Acceptable Use policy. This section does not create a separate enforcement procedure.
Related:Acceptable Use →Terms of Service →
GOVERNANCE & CHANGESNO APPROVED RULE PUBLISHED YET
21. Changes to these AI Terms
Talvrin may update these AI Terms as AI-assisted features, applicable law or Talvrin’s data-flow and risk registries evolve. Material changes are reflected in the version history with an updated effective date.
A public version-history view for these AI Terms is not yet available.
26. RELATED LEGAL / AI PRINCIPLES / TRUST
AI Terms is item seven of nine in the Talvrin Legal sequence.
Each Legal destination is approved individually and in order. Data & Market Information Terms and Legal Notices remain blocked until their own review and approval.
An evidence-first AI boundary, not an AI
marketing page.
Questions about how Talvrin's AI-assisted features work route to Support or the Trust Center below. This page is not a substitute for Talvrin's Terms of Service, Privacy Notice or Data Processing Addendum once fully published.
Research and intelligence platform. No trade execution. No manufactured investment recommendations.