Connectionists: DARPA's GARD: Guaranteeing AI's Robustness against Deception

[Hava Siegelmann]

My friends and colleagues,

Let me introduce

GARD : Guaranteeing AI's Robustness against Deception

 <
https://www.fbo.gov/index?s=opportunity&mode=form&id=be5dbc4deade0d60d52bcbd2e271c521&tab=core&_cview=1
>


26th of February (two weeks from now) - abstracts: not mandatory but
recommended - an opportunity to the researcher to get feedback and to DARPA
to get ready to the proposal
11th of March noon - last time to load proposal - my strong recommendation
is to load by the 10th before midnight to avoid possible losses

The program is 6.1, totally unclassified, no citizenship requirement
And - note the topic is to defend AI from attacks, not to use AI to defend
others.

Please read thoroughly and you can contact me with questions too via
gard at darpa.mil or hava.siegelmann at darpa.mil


All the best -
Hava Siegelmann


PS: I am aware that it is hard to know about funding opportunities with
DARPA. While the grants and contracts are larger than many of the other
agencies, the BAA (call for proposals) is open for 45-60 days only and then
the opportunity is gone - unlike other agencies with regular recurring
dates. If you have ideas how to improve the system, please share it with me
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/connectionists/attachments/20190213/3cea5262/attachment.html>