Dear Readers,
A few days ago I wrote this first line for the Editor's Letter: "Computer security problems have not pervaded the news cycle of late, leading me to believe that in the grand scheme of things, malware is less important than politics and war." But, with the approach of the holidays, malware has picked up its pace, as is apparently a well-worn pattern. The news bag is full, and the need for vigilance is eternal.
I do think that it is worth speculating about a future in which AI takes over our computers and our world. Do we have sufficient safeguards to prevent this? Are there required "manual override" standards for dismantling AI? If the past is any guide, those controls won't be implemented until the problems become serious, and we will be forever patching over problems. Here's to continued research, and research funding, in the new year!
The lineup of accepted papers (89 so far) for next year's Security and Privacy Symposium is an impressive illustration of that spectrum and of the energy of the research community. There are papers about software compartmentalization, neural network security, society's engagement with computer security, cryptographic APIs, "personal cyber insurance", etc., etc. Check out the list, and perhaps you'll be inspired to attend the conference next May in San Francisco.
Epitaph On An AI Computing Platform
Perfection, of a kind, was what it was after,
And the poetry it invented was easy to understand;
It knew human folly like the back of its hand,
And was greatly interested in security;
When it laughed, respectable sysadmins burst with laughter,
And when it cried, the applications got no activity.
Apologies to W. H. Auden