Skip to main content

AI transparency

How we use AI

Clear, honest explanation of what AI does on this platform — and what it does not.

AI on Clap Ideas

AI is used to help people think more clearly and structure ideas better. It is a tool that assists humans — it does not replace human judgment, expertise, or responsibility. Every important action on the platform is performed or approved by a real person.

What AI does

  • Structuring ideas

    When you submit an idea, AI helps turn your raw input into a clear, structured page — title, summary, feasibility tags, research brief, and suggested pathways. You always review and edit before publishing.

  • Detecting duplicates

    AI compares your idea against existing ideas on the platform to find similar or duplicate content. If a strong match is found, you are shown the existing idea so you can add a pathway or comment instead of creating a duplicate.

  • Research briefs

    AI generates a starting research brief covering current technology, known barriers, breakthroughs needed, and suggested experiments. This is a starting point for discussion — not a verified academic document.

  • Ranking and scoring

    AI contributes one component (AI quality score) to the overall idea relevance formula. Community votes, discussion, pathways, and action outputs make up the rest. No idea rises purely on AI score.

  • Moderation

    AI classifies every submitted idea for potential policy violations: spam, harmful content, dangerous instructions, abusive language, and low-effort submissions. AI recommends an action; human admins review and make the final decision on borderline or high-risk cases.

  • Summarisation

    AI-generated content on the platform is labelled clearly: AI-generated, AI-assisted, Under Review, or Speculative. Visitors always know what was produced by AI.

What AI does not do

  • AI does not publish ideas, ban users, move money, or take destructive actions without human approval.
  • AI does not guarantee the accuracy, completeness, or correctness of any research claim.
  • AI does not determine legal ownership, patentability, or feasibility of any idea.
  • AI does not learn from or store your private data beyond what is needed to process your request.

AI labels

Every piece of content that AI helped create or analyse is labelled. Labels include: AI-generated, AI-assisted, Under Review, Speculative, and Requires Breakthrough. These labels are visible on every idea page and card so you always know the provenance of what you are reading.

Verify important information

AI can make mistakes. Before acting on any AI-generated research claim, verify it through independent sources. The platform encourages community members to flag inaccuracies in AI research briefs.

For the full AI usage policy and content moderation policy, see the AI Disclosure and Community Guidelines.

Questions about AI on this platform?

Read the full AI Disclosure document or contact us.