Privacy Policy

Last Updated: November 18, 2026

1. OUR COMMITMENT TO PRIVACY

This Privacy Policy explains how just4o.chat, a service operated by LK Digital Solutions LLC (the "Company"), collects, uses, stores, and protects your data when you use the Service.

OUR CORE PRIVACY PROMISE: We do not sell or rent your personal information, and we do not share your personal information or chat data with third parties for their marketing or advertising. The one exception is the Reddit pixel, which loads only if you opt in to marketing cookies (Section 10.4). We use data to operate, secure, support, and improve the Service, and we disclose data only as described in this Privacy Policy.

For a practical explanation of our moderation, risk, and enforcement approach, see our Safety page.

2. DATA WE COLLECT

We collect the categories of data described in this Privacy Policy to operate, secure, measure, bill, support, and improve the Service. Some of these data are stored on our servers, and some are stored in your browser through cookies, localStorage, or sessionStorage.

2.1 ACCOUNT INFORMATION. When you create an account, we collect your email address and basic profile information through Google Firebase Authentication. This information is used for account management, authentication, and communication about your account. We also collect the age you enter and your confirmation that you are 18 or older.

2.2 USAGE DATA. We collect usage data to manage account limits and billing, including:

  • Model usage counts per user account
  • Daily and monthly usage limits tracking
  • Account tier and subscription status

This usage data is tied to your user account and is used solely for service management and billing purposes.

2.3 SESSION ACTIVITY DATA. When you are logged in, we create session records under your account identifier to support account history, in-product session views, and safety notices. These records may include a session ID, session start time, last active time, last interaction time, active session duration counted in seconds, whether the session is currently paused for inactivity, and the number of chats you sent during that session.

2.4 CONTENT DATA. Your conversations, chat logs, user-generated content, and associated metadata are stored in our backend infrastructure (Google Firestore). We preserve full transcripts so you can retrieve conversations across sessions, devices, and linked accounts, and so automated systems (e.g., memory tools or usage tracking) can operate accurately.

2.5 TECHNICAL DATA. We automatically collect certain technical information when you use our Service, including IP addresses, browser information, and device characteristics. This data is used for security purposes, fraud prevention, abuse detection, rate limiting, and service optimization.

2.6 HUMAN ACCESS LIMITATION. We do not read your conversations. Moderation is automated and can block content or restrict an account without a person viewing it. A person accesses conversation content only if you ask us to, including when you appeal a moderation decision, if we need to in order to fix a technical or security issue, or if the law requires it. Automated systems may process conversation content to deliver features without manual review.

2.7 MODERATION HEURISTIC DATA. When content is blocked and the block is recorded for enforcement, we may store moderation metadata in a separate audit record for enforcement and safety review. That metadata can include your account identifier, the moderation result, timestamps, source, and resulting enforcement state, and may include the request IP address or a derived IP hash. We use these records for enforcement, appeals, abuse investigation, and account safety.

2.7.1 DISTRESS SIGNALS. Signs that you may be in distress are not treated as violations. If a prompt or upload is blocked, we may keep a moderation record of that block as described in Section 2.7.

2.8 CONTACT, REFERRAL, AND ABUSE-PREVENTION LOGS. When you contact us by email or Discord, we receive the message and the contact details you send through that channel. When you submit a referral link, we may store the submission contents, your account identifier, your display name and email address, and the request IP address or a derived IP hash. Contact-form submissions are kept as support records with the email address entered and a derived IP hash. We use these records to review submissions, detect fraud or spam, investigate abuse, and enforce submission rate limits.

2.9 BROWSER STORAGE AND PREFERENCE DATA. We use browser-side storage, including localStorage and sessionStorage, to remember product preferences and temporary state. Depending on the feature you use, this may include theme and font preferences, model-selector filters, voice and TTS settings, sidebar and interface state, recent image-editor state, reference-image selections, project media selections, saved Wikipedia workspace state, session-tracking metadata, session-notice dismissal state, and similar convenience or cache data tied to your browser or account.

2.10 COOKIE AND ATTRIBUTION DATA. We use an essential cookie to store your cookie consent preferences. If you opt in to analytics cookies, Google Analytics may collect page-view and related measurement data, and our first-party product analytics may record limited account milestones such as export parsing, import progress, first continuation of an imported chat, feature adoption, and return days. Those first-party records contain the event category, time, plan, import source, and a coarse success or failure category where relevant. They do not contain prompts, responses, chat titles, file names, memory text, or search text. If you opt in to marketing cookies, the Reddit pixel may collect page-visit and sign-up attribution data. Our servers do not send purchase data or your email to Reddit.

2.11 PAID ACCOUNT IP AUDIT DATA. If you initiate, maintain, or use a paid account, we may store IP addresses observed during checkout, subscription management, or authenticated paid use in a dedicated paid-account IP audit record. We use these records for Stripe-related location checks, fraud prevention, abuse prevention, tax or compliance review, and dispute resolution. This record is included in your data export and is deleted when you delete your account.

2.12 RETIRED PROFILE FIELDS. Older accounts may hold profile fields we no longer use. They are removed when you clear your data or delete your account, and you can ask us to delete them sooner using the contact details in Section 13.

3. HOW WE PROTECT YOUR DATA

We implement security measures to protect your data. Our Service is hosted on Vercel, which provides encryption for data in transit (HTTPS/TLS) and at rest. Data stored in Google Firestore is encrypted using Google's encryption infrastructure.

3.1 ENCRYPTION. Data transmitted to and from our Service is encrypted in transit using HTTPS/TLS. Data stored in Google Firestore is encrypted at rest using Google's encryption standards. We rely on Vercel and Google's encryption infrastructure for data protection.

3.2 ACCESS CONTROLS. Access to user data is limited to what is necessary to provide the Service. We use Google Firebase Authentication for account access control.

3.3 DATA MINIMIZATION. We try to limit collection to what the Service needs.

3.4 APPLICATION SAFEGUARDS. Your stored content is scoped to your account on our servers, and the browser can read only your own records. Uploaded files and generated media are served only to your signed-in account (and, when a request you make needs it, to the AI provider through a short-lived link), and are delivered in a restricted mode that prevents embedded scripts from running. To reduce the risk of content from web pages or files tricking an AI model into leaking your data, images from unrecognized external sites in AI responses load only after you choose to load them. We also apply rate limits and size limits to uploads, imports, and paid features to protect accounts and the Service from abuse.

4. OUR DATA SHARING AND DISTRIBUTION POLICY

4.1 NO SALE OR THIRD-PARTY MARKETING SHARING. We do not sell or rent your personal information. We also do not share your personal information with third parties for their marketing or advertising, except that if you opt in to marketing cookies, the Reddit pixel receives the page-visit and sign-up events described in Section 10.4. You can withdraw that choice at any time in Cookie settings.

4.2 DISCLOSURES NECESSARY TO OPERATE THE SERVICE. We may disclose information in the following circumstances:

  • Legal Requirements: When required by law, court order, or government regulation
  • Account Security: To protect against fraud, security threats, or account abuse
  • Service Providers: To essential service providers who help us host, secure, bill, monitor, analyze, support, or operate the Service, including infrastructure providers and AI model or inference providers. These include Google for infrastructure services; the Vercel AI Gateway, which carries most chat requests and a growing share of image, video, voice, music, and search-embedding requests to the providers named here; AI model developers including OpenAI, xAI, Google AI, Anthropic, Mistral AI (chat models and Voxtral voices), Meta, Microsoft (MAI Image and MAI Voice), Amazon (Nova models, as we add them), Black Forest Labs (FLUX image models), Recraft (image models), and ElevenLabs (voices, music, and sound effects); hosting providers that run models for us, including Microsoft Azure (GPT-4o and search embeddings), Google Vertex AI, Amazon Web Services (Amazon Bedrock), Fal AI (image, video, audio, and other multimodal models, including additional Fal-hosted models we may add over time), Cerebras (GPT OSS 120B Fast and GLM 4.7 Fast) and Fireworks AI (certain older open-weight language models) until those routes are retired, and the inference hosts listed in Section 6.4 for open-weight models (such as DeepSeek, GPT OSS, Kimi, GLM, Qwen, MiniMax, Nemotron, and Thinking Machines Lab Inkling), which we route only to hosts the gateway marks as zero-data-retention; OpenAI, xAI, Microsoft, Mistral AI, and ElevenLabs for voice; and OpenAI for content moderation and TypeSafe AI for safety classification, as described in Section 6A. The privacy handling that applies to each kind of model is described in Sections 6.4 and 6.7
  • Optional Analytics or Marketing Tools: If you opt in to analytics or marketing cookies, limited technical data may be processed by the third-party tools you choose to enable
  • Conversion and Campaign Measurement: If you opt in to marketing cookies, the Reddit pixel may receive page-visit or sign-up conversion events to help us understand ad performance
  • User Consent: When you explicitly consent to specific data sharing
  • Corporate Transactions: In connection with a merger, financing, acquisition, reorganization, bankruptcy, or sale of all or part of our business or assets, subject to applicable legal and confidentiality protections

4.3 THIRD-PARTY SERVICE PROVIDERS. We may engage third-party service providers for essential Service operations. Providers operate under their own terms, and their data practices differ. See Section 6.

5. GOOGLE BACKEND INFRASTRUCTURE

5.1 GOOGLE'S ROLE IN OUR SERVICE. Our Service utilizes Google (Firebase) as its backend infrastructure provider. This includes:

  • Google Firebase Authentication for user login and account management
  • Google Firestore for secure data storage and database operations
  • Google Firebase Admin SDK for server-side processing
  • Google Cloud Platform services for hosting and infrastructure

5.2 USER ACCOUNT MANAGEMENT. Your account authentication and login process is managed entirely by Google Firebase Authentication. This allows you to access your account from any device or location where you can authenticate with Google. Google maintains the authentication infrastructure and handles login security.

5.3 GOOGLE'S PRIVACY PRACTICES. Google's privacy practices and data handling are governed by Google's Privacy Policy and Terms of Service, which apply to the authentication and infrastructure services we use.

6. AI MODEL PROVIDER INTERACTIONS

6.1 OUR ROLE. We build and operate just4o.chat using third-party AI models. Echo, Echo mini, and Echo max are model profiles we configure, and they run on third-party models from our catalog. We also use a gateway model to title chats, a Google model for optional greetings, and an OpenAI embedding model to create search embeddings (served by Microsoft Azure through the Vercel AI Gateway, or by OpenAI directly during the transition described in Section 6.7.1). Our Service sends your requests to AI model providers. For chat, these include OpenAI, xAI, Google AI, Anthropic, Mistral AI, Meta, and certain open-weight models such as DeepSeek, GPT OSS, Kimi, GLM, Qwen, MiniMax, Nemotron, and Thinking Machines Lab Inkling. Most chat models are served through the Vercel AI Gateway with zero data retention requested, and GPT-4o is served by Microsoft Azure, as described in Section 6.4. Cerebras (for GPT OSS 120B Fast and GLM 4.7 Fast inference) and Fireworks AI (for certain older open-weight language models) serve a small number of chat models directly until those routes are retired. For images, video, voice, music, and sound effects, providers may include OpenAI, Google, xAI, Meta, Microsoft, Mistral AI, Black Forest Labs, Recraft, ElevenLabs, and Fal AI (our third-party host for certain image, video, audio, and other multimodal models, including additional Fal-hosted models we may add over time). Depending on the model, these requests may go through the Vercel AI Gateway or directly to the provider, as described in Section 6.7. We also use TypeSafe AI (for safety classification, served through the Vercel AI Gateway, as described in Section 6A), and any successor providers we may add.

6.2 DATA TRANSMISSION TO MODEL PROVIDERS. When you submit queries or prompts through our Service, your input data is transmitted to the AI model provider's servers for processing. This transmission is necessary for the Service to function and provide AI responses. Content is also sent for automated safety checks, whichever model you select, as described in Section 6A.

6.3 MODEL PROVIDER DATA HANDLING. AI model providers have their own privacy policies and data handling practices:

  • Your queries are processed on the model provider's infrastructure
  • Each provider's terms govern how long it keeps data and how it may use it. Some keep data for a limited time for safety and abuse monitoring.
  • The model provider's privacy practices apply to data processed on their servers
  • The handling we are able to request depends on the model you choose. See Section 6.7.

6.4 PROVIDER-SPECIFIC CONSIDERATIONS. Each provider (for example, OpenAI, xAI, Google AI, Anthropic, Mistral AI, Meta, Microsoft, Amazon Web Services, Black Forest Labs, Recraft, Fireworks AI, Fal AI, Cerebras, and ElevenLabs) maintains its own privacy policy and terms. Your use of a given model or voice feature is subject to that provider's terms, including any rights the provider claims to use submitted data for safety, compliance, or product improvement. For clarity, GPT OSS 120B Fast is an OpenAI model served through Cerebras, GLM 4.7 Fast is a Z.ai model served through Cerebras, Fireworks AI is a third-party host for certain older open-source language models until they are retired, and Fal AI is our third-party host for certain image, video, audio, and other multimodal models. Mistral AI models (such as Mistral Large 3, Mistral Medium 3.5, and Ministral 14B) and certain open-weight models (such as DeepSeek, GPT OSS, Kimi, GLM, Qwen, MiniMax, Nemotron, and Thinking Machines Lab Inkling) are accessed through the Vercel AI Gateway. For these models we request zero data retention and restrict routing to inference providers the gateway marks as zero-data-retention compatible, which also excludes using your prompts for training. Depending on the model, the inference provider may be the model developer (for example, Mistral AI) or a hosting provider such as Baseten, DeepInfra, Together AI, Groq, Amazon Bedrock, Nebius, Parasail, Modal, DigitalOcean, Wafer, Morph, Relace, or Runware. We do not route chat requests for these open-weight models to inference hosts headquartered in China. Most OpenAI, xAI Grok, and Google Gemini chat models are also accessed through the Vercel AI Gateway with zero data retention requested and prompt training disallowed, and each request is restricted to a host the gateway marks as zero-data-retention: OpenAI, xAI, or Google Vertex AI, and Microsoft Azure for GPT-4o. We plan to set up chat models we add later, including Amazon Nova models on Amazon Bedrock and newer Google Gemini models on Google Vertex AI, the same way. Anthropic Claude models, and the Google models we use for background tasks such as greetings, memory, and persona generation, are also accessed through the Vercel AI Gateway with zero data retention requested and prompt training disallowed; those requests are served by Anthropic and by Google Vertex AI. If the gateway is unavailable or does not serve the model needed for a background task, that request may go to Google directly under Google's terms. Requests to TypeSafe AI's safety classifier also go through the Vercel AI Gateway, and we send them with zero data retention requested and prompt training disallowed. Zero data retention means the provider does not retain your inputs or the model's outputs after generating a response, and the Vercel AI Gateway itself does not store request or response bodies.

6.5 MEMORY AND BACKGROUND PROCESSING. When automatic memory is on, memory features may send content to model providers other than the one currently answering your chat (for example, to xAI, Google AI, or Anthropic) to evaluate whether a memory should be created, updated, or deleted. That content can include your recent conversation turns and your saved memories. Automatic memory can save personal details you share, subject to the checks in Section 6A.7. You can review and delete memories or turn memory off in the Service. Live assistant responses in the chat window come from the model you selected. Some safety responses use fixed text instead (Section 6A), and AI replies may be automatically adjusted or blocked.

6.6 DATA RETENTION IN OUR SYSTEMS. We retain both user inputs and AI-generated outputs in our Google infrastructure so that you can review prior interactions and so automated features continue to function. Retention generally lasts until you delete the content or your account, subject to Section 8.4 and the legal requirements outlined in Section 9.

6.7 OUR GATEWAY COMMITMENT AND PER-MODEL PRIVACY LEVELS. Our goal is to route every model we offer through the Vercel AI Gateway, using the strongest privacy setting each provider makes available there. We are not there yet. Some models and providers do not offer zero data retention or a no-training setting on the gateway, and some are not on the gateway at all. The handling that applies to your content therefore depends on the model you choose. When a request goes through the gateway, it falls into one of three levels:

  • Zero data retention and no training. We request zero data retention and that your content not be used for training. This level applies to the chat models served through the gateway (Section 6.4), Google Gemini image models (served on Google Vertex AI), xAI Grok image models, Meta Muse Image, Microsoft MAI Image, read-aloud voices from xAI (Grok), Microsoft (MAI Voice), and Mistral AI (Voxtral), and search embeddings (an OpenAI model served by Microsoft Azure).
  • No training. We request that your content not be used for training, but zero data retention is not available, so the provider may keep your prompts, uploads, and outputs for a limited time under its own terms. This level applies to OpenAI GPT Image models and all video models (Google Veo and xAI Grok video), because the gateway's video system does not offer zero data retention.
  • Standard provider privacy. A provider's own terms always apply, at every level. At the two levels above we also ask for a limit on top of those terms. The gateway offers neither limit for these models, so the provider's own terms alone decide how long it keeps your content and whether it may use it, including to train or improve its models. This level applies to Black Forest Labs FLUX image models, Recraft image models, and ElevenLabs (read-aloud voices, music, and sound effects).

When live voice mode is available, it uses an xAI Grok Voice model through the gateway at the first level. Your spoken audio is streamed to the provider in real time. We store the text transcript in your chat and do not store the audio. Transcripts go through the same safety checks as typed messages (Section 6A).

6.7.1 TRANSITION THROUGH JANUARY 14, 2027. We are moving image, video, voice, music, and search-embedding models onto the gateway in stages. Until January 14, 2027, a request to a model we offered before this move may go through the gateway or directly to the provider that served it before (OpenAI, Google, xAI, Fal AI, or ElevenLabs), and a gateway request that fails for a technical reason may be retried automatically on the direct route. When a request goes directly to a provider, the levels above do not apply and that provider's own terms govern (Section 6.3). After January 14, 2027, we plan to retire the direct routes for these models and to remove the models that have no gateway route. If we need to take a chat provider off the gateway temporarily to keep the Service working, requests for that provider's models may go to it directly under its own terms.

6.7.2 SWITCHING MODELS. Because the level depends on the model, switching models can change which companies receive your content, how long they may keep it, and whether they may use it. The model picker shows a privacy level for each model (zero data retention, no training, or standard provider privacy) so you can check before you choose. The level shown describes the gateway route; during the transition in Section 6.7.1 a request may take a direct route instead. Before you switch models, we encourage you to review Vercel's AI Gateway documentation, Privacy Policy, and Acceptable Use Policy, along with the terms, usage policies, and privacy policy of the model's developer and of any hosting provider named for that model, and to consider what the change means for the content you send.

6.7.3 WHAT THESE LEVELS DO AND DO NOT COVER. These levels describe the settings we request and what Vercel and the providers state they offer. We do not operate their systems and cannot independently verify their handling. The levels concern providers only: they do not change what we store in our own systems (Section 6.6), and the safety checks in Section 6A apply whichever model you choose.

6.7.4 WEB SEARCH. Some models search the web with a search tool run by their own provider, under that provider's terms. Other models have no search tool of their own: the open-weight, Mistral AI, Meta Muse Spark, and Amazon Nova chat models served through the Vercel AI Gateway (Section 6.4). For these models, web search is provided by Perplexity, reached through the gateway. When one of these models searches, the search queries it writes are sent to Perplexity, and those queries can include words and details from your conversation. The zero data retention and no-training settings in Section 6.7 apply to the model that answers you. They do not apply to Perplexity. Searches are subject to Perplexity's Terms of Service and Privacy Policy. A search can run when you turn web search on, and also when your message appears to ask for current information.

6A. SAFETY CLASSIFICATION AND CRISIS DETECTION

6A.1 AUTOMATED SAFETY REVIEW. Automated systems, including third-party tools, review messages, prompts, uploads, and AI outputs for safety and crisis signals. These systems may block a message or prompt, and may change or block a reply. Safety Mode, which you can turn on, applies stricter moderation.

6A.2 CRISIS SUPPORT. When a message may signal suicide, self-harm, or an eating disorder, the Service may respond automatically with support resources, including the 988 Suicide & Crisis Lifeline. The reply may be fixed text instead of an AI response. See our Crisis Protocol.

6A.3 NOT A CRISIS SERVICE. These checks are automated. They are not therapy or a crisis service. If you are in crisis, call or text 988 in the US, or contact local emergency services.

6A.4 WHAT WE KEEP. Crisis checks don't add a crisis flag to your account. Support replies are saved in your chat history like other messages until you delete them. Distress signals are handled as described in Section 2.7.1.

6A.5 DAILY COUNTS. We keep daily totals of safety responses. These totals hold no user identifier, message, or score, so they remain after an account is deleted.

6A.6 PROVIDERS FOR THESE CHECKS. These checks use OpenAI's moderation service and TypeSafe AI's classifier, reached through the Vercel AI Gateway, as described in Sections 4.2 and 6.4.

6A.7 MEMORY CHECKS. Memories proposed by Automatic Memory Management or imports are checked before they are saved. Automatic memory is designed not to save sensitive personal information unless you ask it to. Certain memories are not saved for safety reasons, including any that suggest an account holder is not an adult. Memories may still include sensitive information you share. You can view and delete any memory at any time in Account > Memories. See our Health Privacy Notice.

6A.8 PERSONA CHECKS. When you create or change a persona, it is checked. If a persona claims to be a real person, tells the AI to hide that it is an AI, or presents itself as a licensed professional such as a therapist, doctor, or lawyer, those parts are revised automatically so the persona stays clear that it is an AI, and you are told it was adjusted. Personas are told they may stay in character but must never claim to be human, and must say they are an AI if asked.

7. INTERNATIONAL DATA TRANSFERS

7.1 CROSS-BORDER DATA TRANSFERS. Your data may be transferred to and processed in countries other than your own, including countries that may not have the same data protection laws as your country of residence. These transfers occur because:

  • Our infrastructure is hosted on Google's global cloud platform
  • AI model providers operate servers in multiple jurisdictions
  • Our Service is accessible worldwide

Some providers are headquartered outside the United States. For example, Mistral AI is headquartered in France. We do not control where a provider processes a request.

7.2 PROVIDER PRACTICES. Transfers made by our providers are handled under their own terms, and their practices differ. See Section 6.

7.3 YOUR RIGHTS REGARDING TRANSFERS. Please contact us if you would like to know more about where your data is processed. Where applicable law gives you a right to information about international transfers, you can use this route to exercise it.

8. YOUR RIGHTS AND CHOICES

8.1 ACCESS TO YOUR DATA. You have the right to access the personal data we hold about you. You can view and download your chat history, account information, and other stored data through the Service interface.

8.2 DATA PORTABILITY. You can export your data in a portable format at any time from Account settings. The export includes your chats, memories, personas, projects, customization settings, uploaded files, generated images and audio, saved Wikipedia media links, your account record (such as plan, usage, and preference fields), usage and credit-purchase receipts, session history, first-party product analytics records, and any paid-account IP audit record. The export does not include moderation audit records, memory change history and background memory processing records, or the profile fields described in Section 2.12. Records kept for moderation enforcement or abuse prevention may be provided on request where the law requires it.

8.3 DATA CORRECTION. You can update or correct your account information and personal data through the Service settings.

8.4 DATA DELETION. You can delete your account at any time. Account deletion is designed to permanently remove your stored account data (including chats, memories, personas, projects, files, generated media, usage and session records, and analytics records), your sign-in account, moderation audit records and referral submissions tied to your account, any paid-account IP audit record, and contact-form records sent from your account email address, and it cancels active subscriptions. Payment and billing records held by our payment processor, Stripe, are retained by Stripe under its own policies and legal obligations. A non-identifying account identifier may remain in aggregate counters, such as promotion redemption counts, after the account itself is deleted. Daily safety totals (Section 6A.5) also remain as non-personal data.

8.4.1 HOW DELETION WORKS. Deleting a chat or memory removes it from your account. Related records, such as memory change history and search indexes, are removed when you clear your data or delete your account. Clearing your data in Account settings removes your content but keeps your account, including billing, usage, and moderation status fields, and moderation audit records. Deletion removes data from our active systems. Backups and copies held by providers follow their own retention (Section 6.3). We may also keep certain records where the law requires or permits it (Section 9.7).

8.5 OPT-OUT RIGHTS. You can opt out of analytics and marketing cookies through Cookie settings. However, some data collection is necessary for the Service to function.

8.5.1 RESTRICTION AND OBJECTION. Where applicable law gives you these rights, you may ask us to restrict how we process your personal data or object to processing based on our legitimate interests, including product analytics. You can withdraw consent for analytics or marketing cookies at any time in Cookie settings; withdrawing consent does not affect processing that happened before you withdrew it.

8.5.2 HOW TO MAKE A REQUEST. Most requests can be completed directly in Account settings (export, clearing data, or deleting your account). For any other request under this Section, contact us using the details in Section 13. We respond within the time required by law, and may ask you to confirm that you control the account before acting on a request.

8.6 COMPLAINT RIGHTS. If you believe your privacy rights have been violated, you have the right to lodge a complaint with the relevant data protection authority in your jurisdiction.

9. DATA RETENTION

9.1 ACTIVE ACCOUNT DATA. We retain your data for as long as your account is active or as needed to provide the Service. Chat logs, conversation history, and memory artifacts are generally retained until you delete them or close your account, subject to Section 8.4.1.

9.2 SESSION ACTIVITY RECORDS. Session activity records tied to your account, including active duration and chats-sent counts, are retained with your account data so you can view recent sessions and so we can provide safety-related usage notices. These records are deleted when your account data is deleted, subject to legal retention requirements.

9.3 MODERATION AUDIT METADATA. Moderation audit metadata, including your account identifier, categories, scores, timestamps, enforcement outcomes, and any associated IP address metadata or IP hashes, may be retained for safety, abuse prevention, policy enforcement, appeals, and legal compliance. These records do not contain the blocked prompt text itself. We keep them until your account is deleted, unless the law requires us to keep them longer. Clearing your data does not remove them. Daily safety totals are described in Section 6A.5.

9.4 CONTACT, REFERRAL, AND RATE-LIMITING RECORDS. Contact records, referral submission records, and related rate-limiting or abuse-prevention logs may be retained for support, fraud-prevention, anti-spam review, enforcement, and legal compliance.

9.5 PAID-ACCOUNT IP AUDIT RECORDS. Paid-account IP audit records are retained while your account exists, so we can resolve billing disputes, fraud, and compliance questions, and are deleted when your account is deleted, subject to legal retention requirements.

9.6 ACCOUNT CLOSURE. When you close your account, we aim to delete your personal data within a reasonable time, subject to Section 8.4.1 and legal retention requirements.

9.7 LEGAL RETENTION. We may retain certain data longer if required by law, for legal proceedings, or to resolve disputes.

10. COOKIES AND TRACKING

10.1 COOKIE USAGE. We use cookies and similar technologies to enhance your experience and provide the Service. We also use localStorage and sessionStorage for product preferences and temporary workspace state. Cookies are small text files stored on your device. localStorage and sessionStorage are browser storage mechanisms that let the Service remember settings and short-term state.

10.2 ESSENTIAL COOKIES AND ESSENTIAL BROWSER STORAGE. We use:

  • Consent Cookie: We store a cookie named j4o_cookie_consent so the Service can remember whether you accepted or rejected analytics and marketing cookies
  • Preference Storage: We use localStorage or sessionStorage for preferences and convenience features such as theme, font, model filters, voice settings, sidebar state, image-editor state, and other temporary workspace data
  • Session and Safety State: We use sessionStorage for session-tracking metadata and related notice state so the app can maintain recent-session context inside your browser session

10.3 ANALYTICS COOKIES. If you opt in to analytics cookies, we load Google Analytics with IP anonymization enabled. Google Analytics may set cookies such as _ga,_gid, _gat, and _ga_* and may receive page-path and related measurement data when you navigate the site. We may also create first-party, account-scoped product analytics records for a small set of activation and retention milestones. These records are written only by our authenticated backend, cannot be read or changed directly by the browser, exclude user content, and are automatically scheduled for deletion after no more than 400 days.

10.4 MARKETING COOKIES AND ATTRIBUTION. If you opt in to marketing cookies, we load the Reddit pixel, which may set cookies such as rdt_uuid, rdt_cid, and relatedrdt_* cookies and may receive page-visit or sign-up conversion events. If you don't opt in, the pixel doesn't load. Our servers don't report purchases or send your email to Reddit.

10.5 THIRD-PARTY COOKIE CONTROL. If you opt in to analytics or marketing cookies, third-party providers such as Google Analytics and Reddit may set cookies or receive event data subject to their own privacy policies. We do not control those third-party cookies once their tools are enabled. We treat a Global Privacy Control (GPC) signal from your browser as a choice to opt out of marketing cookies, and the Reddit pixel does not load while it is on. Our Service does not respond to browser Do Not Track signals. If you opt in to analytics or marketing cookies, these providers may collect information about your activity over time and across websites.

10.6 FIRST-PARTY SESSION TRACKING. Separately from cookies, logged-in use of the Service may create first-party session records stored with your account in our backend. Those records are used to show your current session time, count chats sent within a session, pause the timer during inactivity, and show break reminders after long sessions. This is product functionality and safety telemetry, not third-party advertising.

10.7 COOKIE MANAGEMENT. You can control cookie settings through your browser and through the in-product Cookie settings. Disabling analytics or marketing cookies may cause us to stop loading the associated client-side tags and to clear certain cookies we can identify, but it may not affect browser storage used for essential product functionality. When a signed-in user withdraws analytics consent, the Service also requests deletion of that account's first-party activation analytics records.

11. CHILDREN'S PRIVACY

Our Service is intended only for adults who are at least eighteen (18) years old. We do not knowingly permit use by minors or knowingly collect personal information from anyone under 18. If we learn a user is under 18, we restrict or close the account and delete data as the law requires.

12. CHANGES TO THIS PRIVACY POLICY

12.1 POLICY UPDATES. We may update this Privacy Policy at any time to reflect changes in our practices, technology, service providers, or legal requirements. Material changes take effect after notice.

12.2 NOTICE OF CHANGES. We will notify you of material changes by:

  • Posting the updated policy on our website
  • Possibly sending an email notification to registered users
  • Providing in-Service notifications

12.3 EFFECTIVE DATE. Changes that are not material take effect when posted. Material changes take effect after we give notice as described in Section 12.2. For material updates, we may require you to review and re-accept the updated policy before continuing to use the Service.

12.4 PROVIDER AND LAW-ENFORCEMENT REQUESTS. We may disclose stored data when required to comply with binding law, lawful process, enforceable governmental requests, or when reasonably necessary to enforce our terms, investigate abuse, protect users or the public, or satisfy contractual obligations tied to integrated providers. Where appropriate, we will disclose only the information we reasonably believe is necessary for that purpose.

13. CONTACT US

If you have any questions about this Privacy Policy or our data practices, please contact us at: just4o.chat@gmail.com

For privacy inquiries or to exercise your rights, please contact us at: just4o.chat@gmail.com

PRIVACY COMMITMENT: We do not sell your personal information, and we never share chat data for marketing. The only marketing tool, the Reddit pixel, runs only if you opt in. Your data is used to operate, secure, support, and improve the Service as described in this Privacy Policy.

NOTE: For more information about our privacy features and user controls, please visit our Privacy page.