# Lee wrote that today’s generative models can render images “virtually indistinguishable” from records of actual abuse…
*Canonical page:* https://truthscore.com/claim/lee-wrote-that-todays-generative-models-can-render-images-virtually--03ddd3c6
*Numbers under lens:* **Evidence-weighted** (`log-odds@1`)
*Machine twins:* https://truthscore.com/claim/lee-wrote-that-todays-generative-models-can-render-images-virtually--03ddd3c6.md · JSON at https://truthscore.com/api/v1
Lee wrote that today’s generative models can render images “virtually indistinguishable” from records of actual abuse, a capability “likely unimaginable” when Free Speech Coalition was decided.

## Voices

*Counted, never weighed.*

- [reddit.com](https://truthscore.com/entity/9758c1c4) — independent

## Sources

- [Seventh Circuit Shields In-Home AI Virtual CSAM Possession While Production, Distribution, and Transfer Charges Stand](https://truthscore.com/source/f6032160)