Legal/compliance question — data extraction methods for YouTube, TikTok, Instagram, and X

Hi,

I have a compliance question about how Supadata's transcript/metadata feature works across the platforms you support, specifically around copyright and each platform's terms of service.

As I understand it, at least for YouTube, the flow is: fetch/download the audio from a video, transcribe it, and return the transcript via the API. I'd like to understand the legal basis for this across YouTube, TikTok, Instagram, and X, since I imagine the underlying method differs by platform. Could you break out the following for each of the four platforms separately?

1. Extraction method per platform: For each of YouTube, TikTok, Instagram, and X, does Supadata retrieve transcripts/metadata by downloading and transcribing audio/video content directly, by using each platform's official API (e.g., YouTube Data API, Instagram Graph API, TikTok API, X API), by using officially-provided captions where they exist, or by some other means? Please specify per platform, since the answer may not be the same across all four.

2. Reproduction of copyrighted content: Where audio/video is downloaded and copied (even temporarily) to produce a transcript, that involves making a copy of copyrighted content. A video being publicly viewable on a platform grants a license to view/stream it through that platform's own interface — it doesn't automatically grant a license to download or copy the underlying media file. Has Supadata obtained a legal opinion on whether this is a permissible reproduction under copyright law (e.g., under a fair use or similar analysis) for each platform where this applies? This is a copyright question independent of whether anyone has signed an agreement with the platform, since copyright protection doesn't depend on a contract.

3. Platform-specific ToS restrictions on automated access:

   • YouTube: The Terms of Service prohibit accessing the Service "using automated means (such as robots, botnets, or scrapers)," with an exception only for public search engines following robots.txt or with YouTube's prior written permission — implying the base rule already covers automated access to public, logged-out content. The YouTube API Services Developer Policies separately prohibit scraping YouTube data/content and prohibit using any technology other than YouTube API Services to access YouTube audiovisual content.

   • Instagram: Meta's Terms of Use prohibit accessing or collecting data "using automated means" without prior permission.

   • TikTok: TikTok's Terms of Service prohibit scraping, crawling, exporting, or extracting data or content "using any automated system or software... except as approved in writing by TikTok."

   • X: X's current terms prohibit crawling or scraping "in any form, for any purpose" without prior written consent, and include a liquidated-damages clause for unauthorized bulk access.

   For each platform where Supadata's method isn't going through the official API, what's the basis for operating outside these restrictions?

4. Indemnification: If a customer's use of this data were challenged on copyright or ToS grounds for any of these four platforms, does Supadata's customer agreement provide indemnification covering how the underlying data was sourced (as opposed to how the customer uses the API)?

We're evaluating whether to build further on this API across multiple platforms and need a clear, platform-by-platform answer before doing so.

Thanks