at://
/
app.bsky.feed.post
/
3mnw5tnogjc2h
sign in
All
4
Record
2
Post
1
PostEmbed
1
Post
by @danabra.mov
PostEmbed
by @danabra.mov
Record
by @jimpick.com
Record
by @atsui.org
+ new component
Post
TLDR: Not reliably enough www.aisi.gov.uk/blog/reality...
17h
www.aisi.gov.uk
A new benchmark grounded in how real users actually probe AI identity during interactions – covering five languages, across text and speech.
RealityTest: Do AI systems disclose their identity when asked? | AISI Work
David Barnard-Wills
