AI Disclosure
Last updated 21 October 2026
Seen & Amplified™ is an AI platform, and you are trusting it with the most personal thing in your business: how you sound. This page says exactly what that involves, in the detail you would want before handing it over.
The four things people actually want to know.
- Generated content is for you to review. An automatic post publishes only after you authorise its schedule or choose to publish it.
- FeiModern does not use your material to train models. Provider handling is explained precisely below.
- Your content is never used to generate anyone else's drafts.
- As between you and FeiModern, you retain your rights in inputs and outputs. AI output may not be unique or copyright-protectable.
01Which AI services we use
We use API services from Anthropic and Google Gemini for writing, analysis and personalisation. Gemini also supports audio transcription and image generation, including workshop-slide images where that feature is available. Image generation may also use Meta or OpenAI. The model used depends on the task.
We do not build or train our own models. Model versions can change. Providers and the purposes for which they receive information are listed in our Privacy Policy.
02What is sent to an AI provider
The system sends relevant task inputs and workspace context to a provider. Depending on the feature, this can include your profile, brand details, prior conversation, saved documents, contact or business activity, imported public material and voice examples. Audio transcription sends the recording to Google; image generation sends its prompt and any selected reference material.
Only submit third-party personal information, images, recordings or confidential material when you have the authority to use it for this purpose. Avoid highly sensitive information unless it is necessary for the feature and you have a lawful basis to provide it.
AI processing can also run in the background as part of a feature you use, such as inferring voice preferences after edits or preparing an enabled scheduled report. It is not limited to pressing a button labelled “Generate”.
Inputs are processed to provide your workspace's features. FeiModern does not use them to train its own AI models or to generate unrelated members' drafts. The feature cannot produce AI output without sending relevant inputs to a provider.
03How Voice Calibration works
Voice Calibration can store a generated draft and your edits, an unchanged draft you accept or schedule, and a draft you discard. It uses those examples to personalise later work within your workspace.
An AI provider may analyse accumulated examples to infer writing preferences. Those preferences are saved and used as instructions in future requests. This is personalisation using examples and instructions, not training a shared model.
Calibration examples and inferred preferences are stored separately from drafts. Deleting a draft does not necessarily erase them. Contact us to request their removal; retention and deletion details are in the Privacy Policy.
04Review, scheduling and publishing
You are responsible for checking generated material before using it. The platform can prepare drafts and suggestions without sending them to your audience.
When you choose automatic social publishing, you authorise the selected content and media to be sent through Post for Me to the accounts and time you choose. Once scheduled, delivery can happen without another approval at the send time. A calendar-only entry or a drafted message is not the same as an automatic publishing instruction.
AI apps you connect yourself. If you connect your own AI app through Connect your AI, that app can prepare a post, but it must show you the exact post and get your yes before it is sent or scheduled. If you set that app to approve actions automatically, it may skip asking you. That setting is yours to control in the app. Information the app reads from your workspace is processed by the company that runs it, under its own terms, not ours. See our Privacy Policy for what a connected app can read and how to disconnect it.
Use the cancellation control for pending automatic posts before disconnecting an account or closing your workspace. Deleting a local draft or disconnecting a service is not a guarantee that an already scheduled or published copy has been removed. You remain responsible for confirming that a cancellation has completed before relying on it.
05What AI gets wrong
Being straight about the limits is part of the disclosure.
- It can be confidently wrong. Facts, figures, dates and quotes need checking before you publish them.
- It does not know what it does not know. It works from what you gave it. Thin input produces generic output.
- It can land near something that already exists. Phrasings recur. If a line matters, check it is yours.
- It is not advice. Nothing it produces is legal, financial, medical or therapeutic guidance, whatever it sounds like.
You are responsible for what you publish under your own name. That is also the point of the product: it gets you to a draft, and your judgement makes it worth reading.
06When AI labels or disclosures are required
Whether you must label AI-assisted content depends on applicable law, the kind of content and the rules of the platform where you publish. Follow any required labels, disclosures or provenance requirements. Editing or approving output does not automatically remove every obligation.
Special requirements may apply to realistic synthetic images, audio or video and certain public-interest content. Do not represent invented events, endorsements or people as real. The European Commission's Article 50 guidance explains relevant EU transparency duties. This page does not certify compliance with every requirement that may apply to the platform or your use.
07Your data and output
| Question | Answer |
|---|---|
| Is it used for training? | FeiModern does not train its own models on your content. Anthropic's commercial API excludes general model training by default; Gemini Paid Services terms depend on the applicable service and billing configuration. Safety monitoring is separate. |
| Do providers retain requests? | Anthropic's standard API retention is within 30 days, with feature, contractual, safety and legal exceptions. Google's Gemini API abuse monitoring specifies 55 days for prompts, context and outputs. Authorised provider staff may review flagged content. Image requests sent to Meta or OpenAI follow those companies' own API terms. |
| Can others access my work? | Access depends on your collaborator/strategist permissions. FeiModern staff have operational access for support, security and requested services. Provider access is described in our Privacy Policy. |
| Is it encrypted? | Data is encrypted in transit and at rest on Google Cloud infrastructure. The service is not end-to-end encrypted: it must process relevant content to run the features. |
| Who owns outputs? | As between you and FeiModern, you retain your rights in inputs and outputs. We cannot guarantee uniqueness, copyright protection or freedom from third-party rights. |
| Can I export or delete information? | Use the exports and deletion controls available for the feature, or email us for access or deletion requests. There is no single self-service export or deletion of all records. Separate calibration records, storage recovery and provider retention can remain relevant. |
See Privacy Policy sections 8 and 12 for retention details and links to current provider policies. Google distinguishes general model improvement from safety and policy-enforcement processing; no-training terms do not mean zero retention.
08Changes to this disclosure
If we add a provider, change how your content is used, or change any of the handling described above, we will update this page and the date at the top, and we will update the sub-processor table in our Privacy Policy before the change goes live. If a change materially affects how your content is handled, we will email you as well.
09Questions
If something here matters to your decision and this page did not answer it, email hello@feimodern.com and ask. A straight answer before you join is better for both of us than a surprise afterwards.