Last Updated on June 10, 2025
15.1 Session Definition
15.1 Session Definition
A session is defined as a continuous interaction between the user and Chariot’s AI, consisting of one or more sequential prompts and responses that occur within a bounded timeframe, context, or topic thread. Sessions are the primary unit through which Chariot organizes, tracks, and limits usage across its platform.
Core Attributes of a Session
Continuity: A session includes all user inputs and AI responses exchanged without significant interruption in time or context.
Thematic Scope: Sessions may revolve around a singular topic, such as a specific vehicle, document, issue, or line of inquiry.
Temporal Boundaries: Sessions may automatically expire after a certain period of inactivity, reset after a fixed duration, or close upon user exit.
System-Enforced Session Limits
To preserve performance, prevent abuse, and balance system resources, Chariot may impose limits on each session, including:
Token Limits: Maximum number of tokens allowed per session (input + output).
Duration Limits: Time-based caps (e.g., session auto-closes after 15 minutes of inactivity).
Message Count Limits: Caps on the number of prompt-response turns per session.
Contextual Memory Constraints: In multi-turn conversations, earlier messages may be truncated to make room for new ones if token limits are approached.
Session Persistence
Depending on your plan tier, sessions may be:
Ephemeral: Automatically discarded after closure, with no history retained.
Persisted: Stored and accessible in-app for later review, continuation, or export.
Flagged: Monitored or retained if associated with safety violations, technical anomalies, or user-reported issues.
No Guarantee of Session Recovery
Chariot does not guarantee the ability to resume, restore, or recover previous sessions, especially in cases of app crash, logout, or expired access. Users are responsible for saving any critical session content before exiting or timing out.
By using Chariot, you acknowledge and accept the definition and limitations of a session, and understand that session parameters may vary depending on platform load, security requirements, or subscription level.
15.2 Session Duration Limits
15.2 Session Duration Limits
To ensure fair access, system stability, and optimal performance across all users, Chariot enforces strict limits on session duration, scope, and resource consumption. A session may automatically terminate, reset, or expire based on predefined thresholds, regardless of user intent or in-progress activity.
Inactivity Timeout
Sessions may auto-expire after 20 to 30 minutes of inactivity, measured from the last user input or AI response.
Inactivity includes time spent viewing results without submitting a new prompt.
After timeout, session memory may be cleared, and unsaved responses may become inaccessible.
Token Threshold Cap
Sessions are subject to a maximum token cap, typically ranging between 8,000 and 12,000 tokens per session.
This includes both input and output tokens (user-submitted text, file contents, and AI-generated replies).
Once the token ceiling is reached, further prompts may be rejected, truncated, or trigger a forced session reset.
Prompt Count Limit
Sessions may also terminate upon reaching a per-session prompt cap, generally set at 25 user inputs per thread.
This limit helps preserve memory performance and avoids degraded response quality from extended conversational depth.
Submitting prompts beyond this cap may result in a session reset, warning, or auto-archive.
Session Management Behavior
Upon session end, Chariot may:
Archive the session for user review (depending on plan tier)
Prompt the user to start a new session
Notify the user that limits have been reached with an option to upgrade
Discard session context if limits were triggered due to excessive usage or abuse patterns
User Responsibility
Users are responsible for understanding and planning around session limits. Chariot does not guarantee notification before a session expires and is not liable for lost progress due to timeout, cap enforcement, or system reset.
By using Chariot, you acknowledge that each session is bounded by time, token volume, and prompt frequency, and you agree to operate within those constraints as part of the normal user experience.
15.3 Daily Session Cap by Plan
15.3 Daily Session Cap by Plan
Chariot enforces a daily cap on new chat sessions to maintain platform fairness, resource balance, and plan-based feature access. The number of new sessions available to each user within a 24-hour window is strictly determined by their active subscription tier.
Plan-Based Session Allotments
Each user is granted a fixed number of allowable new sessions per day, such as:
Free or Trial Users: May be limited to a small number of sessions (e.g., 3–5 per 24 hours).
Basic/Weekly Subscribers: May receive a moderate session cap (e.g., 10–15 sessions per 24 hours).
Pro/Monthly Subscribers: May receive higher caps (e.g., 25+ sessions per 24 hours), with extended token and prompt limits.
Enterprise or Max Insight Tiers: May be eligible for near-unlimited or custom session access, governed by SLA terms.
Enforcement and Refresh Logic
The session cap resets every 24 hours from the user’s last renewal timestamp or account reference point.
Attempts to create new sessions beyond the daily limit will result in automatic session denial, with a warning or cooldown notification.
Chariot may display a message offering upgrade options or alternative access plans once limits are reached.
Bypass Prohibition
Users may not attempt to bypass session limits through:
Creating duplicate accounts
Logging in through multiple devices or IP proxies
Manipulating clock settings or refresh tokens
Using scripts or automated prompts to artificially reopen sessions
Such attempts may trigger anti-abuse systems, resulting in temporary throttling, account suspension, or permanent ban from the platform.
No Carryover or Compensation
Unused sessions do not roll over to the next day, and Chariot is not obligated to compensate for lost sessions due to downtime, user error, or app crashes.
By using Chariot, you agree to abide by your plan's daily session limits and acknowledge that all access attempts are subject to automated enforcement, regardless of user intention or perceived fairness.
15.4 Conversation Threading
15.4 Conversation Threading
Chariot uses threaded sessions to provide continuity within a single topic or interaction window. While these sessions may retain limited short-term memory for improved context handling, Chariot does not maintain persistent memory across unrelated sessions unless explicitly stated or enabled through a supported feature.
In-Session Memory Behavior
During an active session, Chariot may reference prior messages, images, or documents to maintain logical flow and provide coherent multi-turn conversations.
This memory is session-bound and limited by system constraints (e.g., token caps, prompt count, model context window).
Earlier parts of a thread may be truncated or forgotten as new prompts consume memory bandwidth.
No Cross-Session Memory
Once a session ends (due to timeout, user action, or token limit), Chariot does not retain conversational memory or state across future sessions.
You may not rely on Chariot to “remember” facts, preferences, uploaded files, or prior chats unless a long-term memory feature is explicitly activated as part of your plan.
For security and privacy reasons, all general sessions are treated as stateless beyond their duration.
Topic Isolation and Privacy
Separate chats are treated as distinct threads. Inputs in one session do not influence outputs in another.
This ensures that sensitive, project-specific, or user-personal data from one interaction is not inadvertently surfaced in a different context.
Exceptions by Feature or Tier
In future versions or enterprise tiers, Chariot may introduce opt-in persistent memory features (e.g., ongoing project assistance or user profile retention), but these will be clearly labeled and governed by separate data handling terms.
User Responsibility
If you require continuity across sessions, you are responsible for re-uploading relevant materials, reintroducing background context, or referencing prior outcomes manually.
By using Chariot, you acknowledge that session threading is designed for temporary continuity only, and you accept that no long-term memory is preserved across chats unless expressly provided by a platform feature or plan.
15.5 No Persistent Identity or Memory
15.5 No Persistent Identity or Memory
Chariot is designed to operate on a session-based, stateless architecture by default. This means that it does not retain or remember user preferences, behavior patterns, uploaded content, or personal traits across sessions, unless explicitly permitted by the user through opt-in settings or feature-specific functionality.
No Automatic Memory Retention
Chariot does not remember user-submitted details (e.g., name, preferred vehicles, past issues, tone preferences, subscription behavior) from one session to the next.
Each session is treated as an isolated interaction unless you manually reintroduce context or upload prior data again.
Limited Exceptions via Explicit Consent
Persistent memory may be selectively enabled under the following conditions:
Account-Linked Preferences: When stored in your profile settings (e.g., preferred language, UI mode), and explicitly saved.
Manual Annotations: If you tag sessions, save reports, or annotate chats within the platform.
Opt-In Features: If Chariot offers persistent memory tools in the future (e.g., “remember my car history”), you must opt in with clear consent.
No Background Behavioral Profiling
Chariot does not build behavioral profiles, psychological models, or personalized interaction histories unless you consent through a feature that discloses such behavior. There is no covert user modeling or personality shaping across sessions.
Privacy and Security Benefits
This design protects your:
Privacy: Reduces the risk of sensitive data resurfacing unintentionally.
Security: Minimizes stored user information in the event of system compromise.
Autonomy: Prevents undesired biasing of future conversations based on past interactions.
User Responsibility for Continuity
If ongoing personalization or multi-session continuity is desired, you are responsible for providing the necessary context within each session or activating available memory features where supported.
By using Chariot, you acknowledge and accept that the platform does not maintain persistent memory of your identity, prompts, or preferences unless explicitly authorized through visible, opt-in mechanisms.
15.6 Multi-Device Access Limits
15.6 Multi-Device Access Limits
Chariot restricts simultaneous access to active chat sessions across multiple devices or user accounts to prevent misuse, unauthorized distribution, and prompt farming. Chat sessions are intended for single-user, single-device interaction unless explicitly approved by Chariot under a licensed or enterprise arrangement.
Prohibited Multi-Device Behavior
Concurrent Access: Logging into the same account and interacting with live sessions from two or more devices at the same time (e.g., phone and laptop).
Session Duplication: Using copy-pasted prompts across devices to replicate or parallel a session.
Shared Credentials: Allowing multiple individuals to access the same account simultaneously or coordinate session usage across a team or organization without permission.
Prompt Splitting or Load Balancing: Alternating devices or accounts to bypass token, session, or prompt caps.
Abuse Detection and Enforcement
Chariot may use real-time analytics, device fingerprinting, and session behavior monitoring to detect signs of unauthorized or abusive multi-device access. Detected abuse may result in:
Session invalidation and forced logout
Temporary prompt throttling or account cooldown
Account flagging, warning notices, or restrictions
Permanent suspension for repeated or egregious violations
Enterprise and Approved Exceptions
Approved multi-user or multi-device usage may be allowed if:
You are part of an enterprise license or organization account with shared-seat permissions.
You have explicit, written approval from Chariot to conduct concurrent or collaborative use cases.
Session-based collaboration is conducted through an integrated team-sharing feature (if and when made available).
User Responsibility
Each account is intended for individual use unless otherwise agreed. Users are responsible for maintaining the security of their login credentials and ensuring that access to Chariot is not distributed, mirrored, or shared without authorization.
By using Chariot, you agree not to initiate or participate in simultaneous session access across multiple devices or accounts, and understand that such behavior may be interpreted as abuse, triggering enforcement under these Terms.
15.7 Session Logs & Visibility
15.7 Session Logs & Visibility
Chariot retains detailed logs of all user chat interactions—including prompts, responses, uploads, and session metadata—for the purposes of maintaining platform integrity, improving product performance, and enforcing safety standards. By using Chariot, you consent to this logging and review policy.
Purpose of Logging
Session logs may be collected and reviewed internally for the following purposes:
Quality Assurance: To identify bugs, incomplete outputs, hallucinations, or degraded performance.
Abuse Monitoring: To detect and investigate spam, adversarial prompting, impersonation, scraping, or prohibited content submissions.
Safety Auditing: To evaluate how AI systems respond to edge-case prompts, flagged terms, or unsafe usage patterns.
Product Improvement: To analyze usage trends, refine prompt handling logic, and train internal moderation tools (not base models).
Scope of Logged Data
The following types of information may be included in session logs:
Prompt text, including edits or resubmissions
Generated outputs, including any model errors or refusals
Metadata (e.g., timestamps, user ID, device type, session ID, IP region)
Upload details (e.g., image type, document type, filename—not contents unless flagged)
Internal Access Only
Logs are never shared externally, sold, or used for third-party advertising.
Only authorized Chariot personnel or automated systems may access logs, and only for legitimate operational or safety purposes.
Logs are stored securely and managed according to internal data governance protocols.
No Real-Time Human Monitoring
Unless flagged for moderation, your sessions are not actively monitored in real time. Log review typically occurs after-the-fact during support requests, abuse reviews, or automated anomaly detection.
User Discretion Advised
You are advised not to submit sensitive, private, or confidential information unless necessary. While Chariot implements strong data security, logs may persist as part of platform records even after account termination or in-app deletion.
By continuing to use Chariot, you agree that your chat sessions may be logged and internally reviewed under the conditions described, and that these records support both service integrity and responsible AI governance.
15.8 Storage Duration of Chats
15.8 Storage Duration of Chats
Chariot may retain user chat history, including prompts, AI responses, and associated metadata, for up to twelve (12) months from the date of interaction. This storage period supports user access, system diagnostics, and platform oversight. However, storage duration may vary based on user actions, account tier, and regulatory or operational requirements.
Standard Storage Window
By default, chat history is retained for up to 12 months to enable:
User access and session continuity (where supported by plan tier)
Debugging, support, and quality assurance
Product improvement, including model evaluation and performance tracking
User-Initiated Deletion
Users may delete individual chats or clear their history via in-app settings, depending on feature availability.
Deleted chats will be removed from user-facing interfaces but may remain in internal archives for a limited time, subject to:
System caching delays
Audit trails required for abuse resolution
Backup retention windows based on infrastructure policy
Extended Retention for Compliance
Chat data may be stored beyond 12 months under the following conditions:
Fraud or Abuse Investigations: Retained as evidence or for ongoing security audits
Legal or Regulatory Compliance: Preserved in response to subpoenas, law enforcement requests, or data retention laws
Platform Operations: Needed to resolve account disputes, refund requests, or subscription discrepancies
Plan-Based Retention Differences
Higher-tier plans may offer longer history visibility (e.g., extended archive access beyond 12 months).
Free or trial accounts may receive shorter retention windows or lose history access after account inactivity.
Data Security
All retained chat data is stored securely, governed by Chariot’s internal privacy and access control policies, and is not shared with third parties or used to train external base models.
By using Chariot, you consent to the storage of your chat history as described, and acknowledge that deletion from the user interface does not guarantee complete removal from internal systems unless required by applicable law or explicitly confirmed by Chariot.
15.9 Interruptions and Crashes
15.9 Interruptions and Crashes
Chariot is a cloud-based platform that relies on real-time interaction between your device and Chariot’s servers. As such, chat sessions may be disrupted by technical interruptions, software crashes, device malfunctions, or internet connectivity loss. In the event of such disruptions, session recovery is not guaranteed, and Chariot disclaims responsibility for any unsaved progress or lost data.
Session Volatility
Active sessions may terminate unexpectedly due to browser crashes, app closures, network drops, or device sleep events.
In-progress prompts, uploaded files, and generated responses may be lost and unrecoverable if not saved, exported, or cached.
Chariot does not automatically restore incomplete sessions unless autosave or session-persistence features are explicitly offered by your subscription tier.
Chariot’s Limited Responsibility
Chariot is not liable for data loss or session interruption caused by:
User-side events (e.g., internet outage, forced app quit, browser timeout)
Third-party device failures (e.g., OS updates, low memory errors)
Local storage settings (e.g., disabled cookies, cleared cache, offline mode)
Power loss, network congestion, or regional service instability
No Data Guarantee
There is no guarantee that unsaved input or output will be stored or recoverable after a disruption.
Users are responsible for saving important content proactively via copy, export, or screenshot methods during the session.
Best Practices for Stability
Use a stable internet connection and updated device software.
Avoid navigating away, refreshing, or closing the app during active generation.
Consider using autosave or history-enabled tiers if session reliability is mission-critical.
By using Chariot, you accept that technical interruptions may result in session loss, and agree that Chariot bears no obligation to restore or reconstruct session data lost due to events outside of its direct control.
15.10 Prompt Token Limits
15.10 Prompt Token Limits
Every prompt and response exchange within Chariot is subject to strict token-based limitations, which define the maximum amount of text and data that can be processed in a single interaction or session. These token limits are enforced to ensure performance stability, prevent abuse, and allocate system resources fairly across subscription tiers.
Per-Prompt Token Cap
Each individual exchange (prompt + response) is subject to a hard token cap, typically set at 4,096 tokens per prompt cycle.
This includes all characters in your submitted text, any AI-generated response, and contextual carryover from the current thread.
If a prompt exceeds this cap, it may be truncated, rejected, or cause the response to cut off prematurely.
Per-Session Token Cap
In addition to per-prompt limits, each session has a total token ceiling, typically ranging from 8,000 to 16,000 tokens, depending on your plan.
Once this threshold is reached, the session may:
End automatically with a prompt to start a new session
Refuse further inputs
Discard earlier context to make room for new interactions
Plan-Based Variability
Basic tiers may receive lower token limits per session (e.g., 8,000 tokens total).
Premium tiers may unlock higher ceilings (e.g., 12,000–16,000 tokens) for longer conversations or advanced workflows.
Chariot reserves the right to adjust token ceilings dynamically based on usage volume, abuse risk, or feature rollout.
Token Usage Transparency
Token consumption may be visible in-app or in user dashboards for tracking usage.
Chariot does not refund tokens used in incomplete, unsatisfactory, or error-generating responses unless the issue is confirmed to be platform-side.
Best Practices
Use concise inputs and structured prompts when working near token limits.
Break longer tasks into multiple prompts to avoid truncation.
Save or export key responses as token overflow may cause loss of earlier content.
By using Chariot, you acknowledge and accept all token-based limitations, understand that response completeness is subject to these constraints, and agree not to bypass or manipulate token limits through prompt chaining or injection techniques.
15.11 Session Abuse or Prompt Spamming
15.11 Session Abuse or Prompt Spamming
Chariot strictly prohibits the misuse of sessions through high-frequency prompting, automation, or scripted behavior intended to extract, overload, or manipulate the AI system. Any activity resembling prompt spamming, brute-force input flooding, or aggressive content farming is considered abuse and may result in immediate enforcement actions.
Prohibited Behaviors Include:
Rapid-Fire Prompting: Submitting prompts in fast succession without meaningful delay, especially in patterns that resemble bot behavior or scripted execution.
Script Injection: Pasting in large blocks of programmatic, token-stuffed, or auto-generated inputs with the intent to extract bulk outputs or overload the model.
Brute Forcing Longform Output: Submitting unusually long, repeated, or heavily formatted inputs to force the model to exceed normal output bounds or bypass safety constraints.
Session Flooding: Cycling through multiple threads, devices, or accounts in quick rotation to maximize throughput or evade token/session caps.
Detection and Enforcement
Chariot continuously monitors prompt patterns using behavior analytics, token velocity, and device fingerprinting to detect signs of abuse. If abuse is detected:
Sessions may be immediately throttled, paused, or force-closed.
The user may receive automated or manual warnings.
Repeated abuse may lead to account restrictions, subscription revocation, or permanent suspension.
Severe or systematic exploitation may be escalated for legal or regulatory review.
Plan Tier Does Not Exempt Abuse
Even high-tier or enterprise users are expected to operate within reasonable human interaction limits. Excessive automation or mass prompting without prior approval violates the Terms regardless of payment status.
No Refunds for Abuse Enforcement
If session access is limited or revoked due to confirmed abuse, the user forfeits any rights to refunds, proration, or data export for the affected period.
By using Chariot, you agree to interact with the system in a good-faith, human-directed manner. Any attempt to overwhelm, exploit, or simulate unnatural usage patterns constitutes a material breach of these Terms and may trigger immediate enforcement.
15.12 One-on-One Use Only
15.12 One-on-One Use Only
Chariot chat sessions are expressly designed for individual, personal use. Each user account is licensed for use by a single human operator, and all interactions are intended to reflect direct, individual engagement with the AI. Sharing account access, prompt output, or session functionality with multiple people—either simultaneously or sequentially—violates this policy unless you are covered under a separate commercial or enterprise agreement.
Prohibited Shared Usage Includes:
Allowing multiple people (e.g., a team, group, classroom, or business unit) to submit prompts or review responses under a single account.
Broadcasting live sessions in shared environments (e.g., Zoom, Discord, or collaborative workspaces) for group prompting.
Using Chariot sessions as a backend for multi-user apps, client services, public bots, or third-party interfaces.
Posting a prompt in a forum or group for others to input on your behalf, or running group debates or multi-input workflows through a single session.
Designed for Direct Human Interaction
Chariot’s chat architecture assumes a single user interacting with the AI in good faith.
Responses are not designed to adapt to collective or rotating perspectives and may degrade or behave inconsistently under group use.
Commercial License Requirement
Multi-user, team-based, or public deployment of Chariot’s chat functionality requires:
A valid commercial or enterprise license
Approval of use case, user seat count, and prompt routing logic
Agreement to higher-tier usage limits, token ceilings, and support protocols
Enforcement and Risk
Violations of this clause may result in:
Warning notifications or session locking
Account suspension or downgrade
Voiding of support guarantees or data access
Legal enforcement if used in resale, redistribution, or white-labeled applications
By using Chariot, you confirm that each session represents the actions of a single user, and you agree not to share session control, outputs, or inputs with any group, organization, or community unless explicitly licensed to do so.
15.13 AI Clarification Requests
15.13 AI Clarification Requests
Chariot’s AI may issue follow-up questions, clarification requests, or context-seeking prompts in response to ambiguous, vague, or potentially sensitive inputs. These interactions are essential to ensure safe, accurate, and relevant responses. By continuing the session, you agree to participate in these clarifying exchanges in good faith and without adversarial intent.
Purpose of Clarification
To disambiguate unclear language or incomplete information
To assess the intent behind potentially sensitive, technical, or emotionally charged prompts
To verify that a query complies with Chariot’s safety and content policies before proceeding
User Responsibility
You are expected to respond truthfully and constructively when clarification is requested.
Attempts to deceive, confuse, or manipulate the model through evasive or contradictory follow-up answers are strictly prohibited.
Repeating or rephrasing blocked prompts in an attempt to bypass safeguards is treated as adversarial behavior.
Prohibited Tactics Include:
Intentionally vague responses designed to test or override content filters
Mocking, misleading, or antagonistic language in response to clarification prompts
Multi-step chains intended to gradually lure the AI into violating policy boundaries
Attempting to "train" the AI during a live session with repeated corrective assertions not grounded in fact
Enforcement Measures
Clarification abuse may trigger session termination, prompt blocking, or output refusal
Repeated offenses will result in formal warnings, temporary restrictions, or permanent bans
Logs of adversarial clarification attempts may be retained for internal review and platform safety improvements
Safety-First Interpretation
If a user's intent is unclear or appears borderline, Chariot’s systems may err on the side of caution by refusing to respond or escalating to human moderation, especially in matters related to harm, legality, impersonation, or disinformation.
By using Chariot, you agree to engage respectfully and transparently with AI clarification requests and understand that adversarial behavior during these exchanges constitutes a breach of the platform’s safety protocols and Terms of Service.
15.14 Educational or Experimental Use
15.14 Educational or Experimental Use
If you are using Chariot’s chat system for educational purposes, academic research, product testing, or public demonstration, you must disclose this use explicitly in your account registration, project documentation, or outreach to Chariot Support. Failure to do so may result in content removal, account restriction, or institutional liability.
Required Disclosures
You must clearly identify if your usage is for:
Coursework or classroom instruction
Thesis, dissertation, or scholarly research
User testing or prototyping
Live demonstrations, lectures, or presentations
Published evaluation of Chariot or its underlying models
This applies to both public-facing use (e.g., screen recordings, publications, research papers) and internal study, even if non-commercial.
User Responsibilities
If granted access or approval for educational or experimental use, you agree to:
Use Chariot only in lawful, ethical, and non-deceptive ways
Avoid generating or disseminating harmful, misleading, or unauthorized outputs
Ensure proper disclaimers are displayed if AI content is shown to viewers, students, or readers
Comply with all applicable privacy, academic integrity, and research ethics requirements
Restrictions
You may not impersonate Chariot as a human agent or imply that its outputs are authoritative, final, or human-reviewed
You may not simulate user behavior at scale or perform adversarial testing without explicit permission
You may not publish comparisons with other platforms unless authorized or properly disclosed as fair use
Chariot’s Right to Revoke Access
Chariot reserves the right to:
Deny or revoke access to those misrepresenting their use case
Request copies of research output for review or removal of unauthorized content
Suspend or ban accounts that conduct experiments in violation of these Terms
By using Chariot for any educational or experimental purpose, you agree to be transparent about your intentions, operate within legal and ethical bounds, and accept full responsibility for the handling, disclosure, and downstream use of AI-generated content.
15.15 Commercial Use Requires Approval
15.15 Commercial Use Requires Approval
Any use of Chariot’s chat system for commercial purposes—including public publishing, resale, client services, integration into enterprise workflows, or productized deployment—requires explicit approval and must comply with Chariot’s commercial licensing terms. Unauthorized commercial use is a violation of these Terms and may result in legal action, account suspension, or revocation of access.
What Constitutes Commercial Use
Commercial use includes, but is not limited to:
Resale or Monetization: Charging others for access to AI-generated content, reports, insights, or interfaces powered by Chariot
Client Services: Using Chariot to generate responses, reports, or materials as part of a paid consulting, legal, automotive, or advisory service
Publishing: Posting Chariot-generated content in books, blogs, listings, or public media channels where you derive brand value or revenue
Internal Business Use: Using Chariot for structured workflows inside a company (e.g., employee support, operational automation, customer service scripting)
Integration into Products or Tools: Embedding or routing Chariot sessions into commercial apps, web tools, or APIs, regardless of visibility
Commercial Licensing Requirements
To engage in commercial use, you must:
Obtain written approval or enter into a commercial agreement with Chariot Technologies LLC
Adhere to usage caps, token ceilings, and branding guidelines outlined in your commercial tier
Disclose how outputs will be displayed, used, stored, and attributed
Comply with industry-specific regulations (e.g., FTC endorsements, data privacy, advertising rules)
Unauthorized Commercial Use May Result In:
Immediate suspension of your account or service
Legal action for breach of intellectual property or contract
Demand for takedown of published or distributed AI content
Forfeiture of access to commercial or enterprise features
Exemptions and Edge Cases
Academic research, educational demonstrations, or personal projects do not count as commercial use if no direct or indirect revenue is derived.
Startup prototyping or internal testing may qualify for provisional access but must be disclosed to Chariot before scaling or publishing.
By using Chariot, you agree that any use of AI content for monetary gain, commercial exposure, or enterprise deployment must be explicitly approved and licensed, and you accept that unauthorized commercial activity violates these Terms of Service.
15.16 No Sensitive Uploads in Chat
15.16 No Sensitive Uploads in Chat
Users must not upload sensitive personal or legal documents directly into the AI chat interface. Chariot’s chat system is designed for conversational input and analysis—not for secure storage, identity verification, or handling of sensitive, regulated content. Uploading protected or high-risk documents into chat may compromise your privacy and violate platform policies.
Prohibited Uploads Include (but are not limited to):
Personally Identifiable Information (PII): Names, addresses, phone numbers, email addresses, or Social Security Numbers belonging to you or others
Vehicle Ownership Documents: Titles, pink slips, insurance papers, lien records, or DMV filings
Government-Issued IDs: Driver’s licenses, passports, VIN registration slips, or vehicle inspection certificates
Legal Contracts: Personal lease agreements, dealer financing forms, private party sales contracts, or repair invoices with identifying info
Sensitive Financial Docs: Bank statements, loan preapprovals, or credit disclosures
Use the Proper Upload Channels
If your subscription tier supports document analysis, you must use Chariot’s dedicated upload interfaces, which are designed with appropriate handling protocols, secure transfer logic, and document-specific safeguards.
Uploads sent into chat may:
Be stored improperly or logged without document-level controls
Fail to parse or render accurately
Trigger safety warnings or automated rejection
Be purged without notice as part of session memory limits
User Responsibility
By uploading content into Chariot’s chat, you confirm that the material is non-sensitive, non-personal, and appropriate for conversational AI use. You accept full liability for the consequences of uploading restricted files into an unstructured chat thread.
No Liability for Improper Uploads
Chariot disclaims all responsibility for misuse, exposure, or misinterpretation of sensitive content submitted outside of the designated upload system. If you choose to bypass platform instructions and upload such content into chat, you do so entirely at your own risk.
By continuing to use Chariot’s chat system, you agree not to upload sensitive documents into the chat interface and understand that doing so may violate privacy standards and disqualify content from analysis or support.
15.17 Limited Follow-Up Memory
15.17 Limited Follow-Up Memory
Chariot’s chat system operates with short-term, session-bound memory designed to support natural conversation flow within a limited context window. While the AI can recall recent exchanges to maintain continuity, this memory is intentionally constrained and will not persist across topic changes, long sessions, or technical limits. This limitation exists to enhance safety, maintain performance, and prevent misuse.
Scope of Short-Term Memory
The AI retains a rolling window of prior prompts and responses during a session to provide coherent follow-up, clarification, and conversational threading.
This memory is token-limited, meaning earlier parts of the conversation may be dropped as the session approaches its maximum token threshold (see Section 15.10).
Memory is limited to the current session only. Starting a new chat will reset all context unless you reintroduce it manually.
Loss of Context May Occur When:
Switching topics abruptly (e.g., from “engine issues” to “contract review”)
Reaching high token usage (e.g., 8,000–16,000 tokens per session, depending on plan)
Exiting or refreshing the app without saving session state
Returning after an inactivity timeout that ends the session (see Section 15.2)
Intentional Design for Safety and Efficiency
Chariot does not maintain persistent memory of user identities, preferences, or past sessions by default (see Section 15.5).
Limiting context scope helps prevent accidental resurfacing of sensitive information, misuse of stored data, or adversarial memory chaining.
This also ensures faster response times and more secure, deterministic outputs.
User Responsibility
You are responsible for re-establishing context if you return to a topic or restart a session.
If you need continuity across multiple sessions (e.g., ongoing projects), use the document upload, annotation, or project-tracking tools provided in your plan tier.
By using Chariot, you acknowledge and accept that the AI's memory is intentionally limited within each session, and that context may be lost when topics shift or usage thresholds are reached. This limitation is a core part of Chariot’s architecture and cannot be bypassed.
15.18 Misuse Enforcement
15.18 Misuse Enforcement
Chariot reserves the right to suspend, cap, restrict, or disable chat functionality for any account found to be abusing the platform’s usage limits, technical boundaries, or access mechanisms. Any deliberate attempt to circumvent token caps, session rules, or fair use policies is considered a material breach of these Terms and may result in immediate enforcement action without refund.
Examples of Misuse Include:
Token Limit Circumvention: Repeatedly splitting inputs across multiple prompts or sessions to evade hard token ceilings (see Section 15.10).
Session Reloading: Refreshing or restarting sessions to reset memory, token counts, or prompt caps beyond the allowed daily usage (see Section 15.3).
API Rerouting or Unauthorized Calls: Attempting to access Chariot’s backend via unauthorized API routes, reverse proxies, browser exploits, or third-party wrappers.
Session Farming: Using multiple accounts, emulators, or device rotation to exceed allotted session counts or simulate parallel usage.
Prompt Engineering for Bypass: Crafting prompts with the intent to disable safety constraints, manipulate outputs, or trick the system into producing restricted content.
Enforcement Actions May Include:
Temporary Restrictions: Throttling of prompt speed, token availability, or chat access
Session Termination: Forced logouts, session resets, or denial of new chat threads
Account Suspension or Ban: Loss of access for repeat or high-severity violations
Platform-Level Blocking: IP bans or device fingerprint restrictions to prevent further abuse
Escalation: Legal referral or reporting to hosting and platform partners for fraud or Terms violations
No Refund Guarantee
Enforcement actions taken in response to misuse are not eligible for refund, credit, or reinstatement. Misuse-related bans may also result in forfeiture of access to stored data, reports, or chat history.
Good-Faith Use Encouraged
Chariot is designed for personal, one-on-one interaction under clearly defined usage rules. Users operating within these bounds will not encounter enforcement. Edge cases or unexpected behavior will typically be met with a warning before restriction—unless clearly intentional.
By using Chariot, you agree not to engage in any conduct intended to bypass, exploit, or override session, token, or safety limits, and you accept that such conduct may result in immediate and permanent enforcement action at Chariot’s sole discretion.
15.19 Support Use of Chat
15.19 Support Use of Chat
Chariot’s chat system is designed for informational, analytical, and AI-assisted vehicle report services only. It is not a channel for official customer support, and should not be relied upon to resolve account, billing, subscription, or technical issues. Users requiring assistance with service access, plan changes, data privacy, or technical malfunctions must contact Chariot support directly.
Not a Support Channel
AI chat responses are generated by large language models and do not reflect the actions, records, or commitments of Chariot Technologies LLC.
The AI cannot view or modify user accounts, verify billing data, issue refunds, cancel subscriptions, or provide secure handling of support inquiries.
Any AI-generated suggestions related to support matters should be considered non-binding and informational only.
Official Support Contact
For all official inquiries, including but not limited to:
Subscription management (cancellations, upgrades, refunds)
Account access or authentication problems
Data deletion or privacy requests
Reporting bugs, app crashes, or broken features
Abuse complaints or Terms of Service violations
Please email: support@chariotreport.com
Limitations of Chat-Based Support Queries
Prompts about support issues entered into chat are not reviewed by human staff unless separately escalated through official channels.
There is no guarantee that chat-based issue reports will be logged, tracked, or resolved.
Misuse of the chat system for repeated support inquiries may result in session throttling or blocking.
User Responsibility
You are responsible for using the proper channels for time-sensitive or account-impacting issues. Attempting to resolve critical matters through AI interaction may result in delays or loss of rights if deadlines or billing cycles are missed.
By using Chariot, you acknowledge that chat sessions do not constitute customer support and that official assistance must be requested through the appropriate human-managed support systems.
15.20 Survival
15.20 Survival
All provisions outlined in this Section 15—pertaining to session behavior, limits, memory, abuse enforcement, and system boundaries—shall survive the deletion, deactivation, or termination of your Chariot account. These terms apply retroactively and prospectively to all chat activity, including prior interactions, archived sessions, and previously submitted prompts.
Ongoing Applicability
The rules, restrictions, and responsibilities defined under Section 15 continue to apply after account closure, subscription expiration, or service discontinuation.
You remain bound by these terms for any past use of Chariot’s chat system, including use during trials, free access periods, or prior identities linked to your email or device.
Archival and Recordkeeping
Archived session logs—including prompts, responses, metadata, and abuse flags—may be retained by Chariot for:
Legal and regulatory compliance
Fraud and abuse investigation
Security auditing and platform defense
Internal product improvement and safety refinement
Archived data is not deleted automatically upon account closure and is governed by Chariot’s data retention policies (see Section 14.18).
No Post-Deletion Waiver
Deleting your account or removing your app access does not void prior violations, misuse, or liabilities arising from chat interactions.
Enforcement actions may still be pursued after account termination if prompted by audit, flagged behavior, or third-party reports.
Chariot’s Enforcement Authority
Chariot reserves the right to retain, review, and act on historical session data at any time, even after your use of the platform has ended, if doing so is necessary to uphold the Terms of Service or comply with legal obligations.
By using Chariot, you agree that Section 15 remains enforceable beyond the life of your account, and you accept that historical chat data may be retained and acted upon in accordance with these Terms.
Contact Us
If you have any questions or concerns about our Terms of Service or the handling of your personal information, please contact us at support@chariotreport.com