AI Content Policy

• Text: brand analysis, market and competitor notes, monthly plans, captions, hashtags, replies from the assistant (Pip) and report summaries are written by large language models from OpenAI, with Anthropic's models as a backup.
• Images and video: AI images, image edits, image-to-video and video scenes are produced by Viralance, a media production service run by GSAM Solution LLC, using its AI model providers.
• Voice notes: voice notes you record can be turned into text by an OpenAI speech-to-text model.

AI does not make decisions about you that have legal or similarly significant effects. People at Feedhand may look at your brand and content to support you or to investigate a problem.

Each brand chooses how much AI may be used for visuals: real media only, assisted (the default) or AI allowed. Your own clips and photos are used first; AI motion is added to your own photos next; scenes generated from text come last, and never under "real media only".

Whatever the setting, the Service instructs every visual to follow these rules: your product, packaging, label, logo and colours are never redrawn or changed; a product is not shown if there is no photo of it; synthetic people, faces or hands are never presented as customers, employees or reviewers; and there are no fake reviews, ratings or before-and-after results.

AI can still get things wrong. These rules reduce the risk; they do not replace your review.

Every visual carries a badge in the app: your media, edited with AI or AI-generated. A post that contains an AI-generated or AI-edited visual is marked for an AI label, and the hand-off page tells you how to add it on each platform (for example the AI label setting on Instagram and TikTok, or a line of text on X and Reddit).

Keep the labels and disclose where required. Under Article 50 of the EU Artificial Intelligence Act, from 2 August 2026 anyone who publishes realistic AI-generated or manipulated images, audio or video ("deepfakes") must disclose that they are artificial, and AI-generated text published to inform the public about matters of public interest must be disclosed unless it was reviewed by a person who takes editorial responsibility. Platforms such as Meta, TikTok and X also require or apply AI labels. Other countries may have similar rules.

When you download a post, media files are delivered as produced, without removing embedded provenance information (such as C2PA content credentials or metadata) that the generator may have added. Do not remove it.

Advertising and partnership labels are a separate duty: see section 5 of the Terms of Service.

• Facts, prices, dates, opening hours, product names and anything about health, safety, money or the law.
• That images and videos show your real product correctly and do not mislead.
• That nobody real is shown, imitated or named without permission.
• That the post carries the AI, advertising or partnership labels it needs, and suits the platform and the country where it will be seen.
• That it sounds like you. You can edit, reject or rewrite any draft.

You own the Output made for your brand as section 7 of the Terms of Service describes. AI can produce similar content for others, and purely AI-generated material may not be protected by copyright in some countries.

Do not ask the Service to imitate a living artist's distinctive style, a competitor's protected design or campaign, a celebrity's likeness or voice, or copyrighted characters or music.

We do not use your material or Output to train AI models. Your brand's own preferences (for example the rules you set, the voice examples you choose and what you approve or reject) are used to improve the drafts for your brand only.

Where content is sent to AI providers, it is sent only to produce your Output, under their business terms. See the Privacy Policy.

If the Service produced something harmful, misleading or infringing, reject it and tell us at support@feedhand.com. We use these reports to improve our rules and filters.