Who we are and what this policy covers
LIGHT SPEED FUTURE LTD operates the Light Speed Future ChatGPT plugin and handles its support and privacy enquiries. We are registered in England and Wales under company number 17498684. Our registered office is 82A James Carter Road, Mildenhall, Suffolk, IP28 7DE, United Kingdom. Contact us at [email protected].
This policy covers the plugin, account connections, plugin task records and related support. It does not replace OpenAI’s privacy information for ChatGPT or the terms and privacy information applicable to a separate LSF API account. Website sales enquiries, advertising and API features outside the plugin have their own applicable notices.
We determine how personal data is used for plugin access, security and support. When we process content solely on a business customer’s instructions, the applicable customer and data-processing arrangements govern that processing. The customer is responsible for its instructions and for having a lawful reason to provide personal data in the content.
LIGHT SPEED FUTURE LIMITED in Hong Kong remains responsible for existing LSF API account contracts, balances and settlement. The UK company coordinates relevant requests with that company. Connecting an account does not transfer its contract or balance.
Information we process
We receive information you provide, information supplied through the connected host and identity service, and information returned by LSF services and their providers.
Connection and account information includes identity and account identifiers, authorisation status and scopes, enabled products, account bindings, and relevant allowance or pricing information.
Task inputs include the task-specific prompt, settings, source media URL, content made available through that URL, and confirmations or rights acknowledgements. A URL may include confidential, temporary access parameters.
Task and usage records include draft and task references, timestamps, request settings, estimates, confirmation receipts, status, usage, sanitised errors, and information needed to recover the original task or investigate a charge.
Source checks and results include media properties, a content fingerprint used to validate a source, and temporary result or preview links returned for a task. Support information includes your email address, messages, attachments you choose to send and our related correspondence.
Technical and security information can include IP addresses, request times, authentication events and diagnostic errors in the relevant infrastructure, where needed to operate and protect the service.
Information you choose to share
The plugin does not request your entire chat history. Information included in a tool request may reach us, so provide only what the task needs. Do not submit passwords, API keys, one-time codes, payment-card details, government identifiers, protected health information or other sensitive personal information. Do not provide another person’s personal data without the necessary authority and lawful basis.
Why we use information
We use information to authenticate and authorise access; show enabled products and relevant allowance; prepare and submit tasks you confirm; validate media sources; return and recover results; reconcile usage; provide support; prevent misuse; and meet applicable legal obligations.
Where data-protection law requires a lawful basis, we rely on performance of a contract for services requested by an individual customer, and on legitimate interests where appropriate for operating and securing the service and supporting business customers. These interests include preventing unauthorised access, avoiding duplicate tasks, investigating failures and handling disputes. We rely on legal obligations where a law requires processing. Where consent is required, we explain it separately and allow withdrawal.
Task confirmation authorises the requested service action. It is not blanket consent to unrelated personal-data processing. Without necessary connection and task information, we may be unable to connect your account or perform the requested function.
We do not use plugin inputs to build advertising profiles or train our own general-purpose models. This statement does not make a training or retention promise on behalf of OpenAI or another model or infrastructure provider.
Who receives information
OpenAI and the ChatGPT host environment receive tool results and display task cards under their own arrangements with you. Auth0 provides identity and authentication functions. Cloudflare and hosting, storage, backup, communications and security providers support delivery and operation of the service.
LIGHT SPEED FUTURE LIMITED in Hong Kong and the relevant LSF API, model and video-enhancement providers receive the information needed to execute tasks, return results and administer existing API accounts. This can include prompts, media, source URLs and task settings. BytePlus services participate where used for your selected product.
The storage service hosting your source or output participates when media is accessed. The source URL you supply must be authorised for this use. Authorised staff and service providers may access necessary records for support, security, maintenance and legal obligations. Professional advisers and authorities may receive information where necessary and lawful.
We limit access and disclosures to their purpose. We do not sell plugin personal data. Provider systems may have separate contractual retention and data-use conditions; our handling of media does not mean every recipient keeps no copy.
How long we keep plugin records
The following schedule applies to records controlled by the LSF plugin. We delete or irreversibly anonymise records when their retention period ends, subject to the limited exceptions below. Pseudonymisation alone is not treated as deletion or irreversible anonymisation.
| Record | Retention |
|---|---|
| Connection and product bindings | While enabled, then up to 30 days after disconnection or revocation, except information required for unresolved tasks or legal obligations. |
| Unsubmitted drafts, prompts and source records | Up to 30 days after the last meaningful user activity on the draft or source. Background checks do not restart this period. |
| Submitted task content and source records | While the task or its settlement is unresolved, then up to 30 days after both are resolved. |
| Minimal task, usage and confirmation records | Up to 90 days after the task and settlement are resolved. These records exclude prompt and media content and complete signed URLs. |
| Plugin-controlled operational and security logs | Up to 30 days, unless a relevant incident requires a documented longer hold. |
| Support correspondence and attachments | Up to 12 months after the request is closed. |
| Routine backups of plugin-controlled records | A rolling maximum of 30 days from backup creation. |
A completed task is not necessarily a settled task. We retain information necessary to resolve uncertain acceptance, prevent duplicate submissions, complete settlement and handle an active issue. Longer retention for a specific legal obligation, active dispute or security investigation is limited to the necessary information and purpose, is reviewed, and ends when the relevant need ends.
Financial records held for an existing API account are subject to the Hong Kong operator’s applicable contractual and legal requirements. Copies used by the API service for fulfilment, accounting or support are assessed under their applicable purpose and rules. We coordinate requests concerning those copies; transferring data to the API service is not a way to avoid this policy.
These periods do not describe ChatGPT conversation retention or all independent provider systems. They do not make result links permanent. Source-validation expiry is an operational limit and does not itself delete a record.
Media and backup handling
The plugin’s source-checking process handles media in memory rather than maintaining an ordinary customer media-file archive. It still stores relevant prompts, URLs, source-validation and task records. Temporary output links are provided for delivery; save results you need to storage you control.
Removal from live systems may precede expiry of the corresponding backup. Residual backup copies are restricted from ordinary use and expire under the backup schedule. Before restored data is returned to service, we reapply applicable deletions and retention decisions and verify that deleted records are not reintroduced. Necessary evidence of a request or legal hold is kept separately within the applicable support, security or legal retention rule.
Provider retention
When a task is processed through BytePlus ModelArk, its published data-processing information states that inputs and outputs flagged by its content filter are retained for 180 days in Malaysia. This is a provider content-safety rule, not the retention period for all tasks or all provider records. Other provider retention depends on the service and applicable terms.
We coordinate applicable privacy requests with the responsible API operator and providers. We do not promise that every recipient erases every copy within the plugin’s retention period, or that expiry of a download link proves deletion of underlying content, logs and backups.
International processing
Plugin data may be processed outside your country and outside the United Kingdom. The service involves the Hong Kong API operator and international providers. Plugin hosting includes Malaysia. BytePlus ModelArk’s published information identifies Malaysia, Indonesia and the EU/EEA as possible processing locations. Identity, security, communications and backup providers may use other locations relevant to the selected service. A UK operator does not mean all data remains in the UK.
Where UK or other applicable data-transfer restrictions apply, an applicable lawful transfer mechanism is required. The arrangements depend on the recipient and service and may include adequacy regulations or approved contractual safeguards with the required assessment and additional measures. Contact [email protected] for information about the arrangements applicable to your service and a copy of relevant safeguards, subject to necessary redactions.
Security
We use access controls, account-bound permissions, transport protection and encryption for sensitive stored payloads. Encryption does not mean every database field is encrypted or that authorised operators can never access information. No system can guarantee absolute security. Keep credentials private and use approved media links with access limited to the requested workflow.
Your choices and rights
You can choose what to submit, review a draft before confirmation, stop using the plugin and disconnect or request revocation of its access. Disconnection does not delete earlier records, cancel an upstream task already accepted, or erase your ChatGPT history.
Depending on applicable law and the basis for processing, you may request access, correction, erasure, restriction or portability of your personal data. You may object to processing based on legitimate interests by explaining your situation. Where we rely on consent, you can withdraw it without affecting the lawfulness of earlier processing.
Email [email protected] and describe your request. No special form or subject line is required. An account or task reference can help if available. Do not send passwords, API keys or identity documents initially. We request proportionate verification only where necessary to confirm identity and authority.
We handle requests within applicable legal time limits and explain any justified limitation, extension or need to coordinate with another responsible organisation. Deletion is not an absolute right: we may retain limited information where a legal obligation, active dispute or other applicable exception requires it. We explain the scope and basis of any relevant retention.
If a business customer controls personal data in content, we may direct the request to that customer and assist it under the applicable processing arrangements. You may complain to the UK Information Commissioner’s Office at https://ico.org.uk/make-a-complaint/ or another competent data-protection authority. Contacting us first is helpful but does not remove your right to complain.
Automated processing and children
The plugin uses automated generation, enhancement, validation and access checks to provide its features. We do not use it to make solely automated decisions about people with legal or similarly significant effects. Do not use the service for that purpose.
The service is not directed to children under 13. Users must meet the eligibility requirements of ChatGPT and their LSF account. If you believe a child has provided personal data improperly, contact us.
Changes and contact
We update this policy when our practices change and display the version, effective date and latest update date on this page. We give appropriate notice of material changes before applying a new use where required.
Privacy contact: [email protected]. Postal address: LIGHT SPEED FUTURE LTD, 82A James Carter Road, Mildenhall, Suffolk, IP28 7DE, United Kingdom.