I am working through the compliance for the OSA at the moment
While the majority of the work seems to be in documenting that the organisation has considered and attempted to mitigate the risk of Illegal Harms, I wondered if anyone out there has developed any shareable documentation or automated tooling to help provide an evidence base for a lack of nefarious activity within their forum?
I was thinking possibly:
- Data Explorer queries searching for specific keywords or patterns of usage?
- Lists of Watched Words that could improve flagging of posts and private messages?
- Discourse AI configurations or prompts which might help to flag concerning activity?
- Templates for an improved ‘UK OSA Approved’ Terms of Service?
- Other ‘sysadmin tips and tricks’ for maximising the effectiveness of the protections, with the minimum of additional work?
I am just as concerned as anyone else would be about the privacy implications of all this, but if audited we would need to be able to show we invested an appropriate amount of effort checking that nothing Illegal was going on, and can present (and review) this evidence annually.