Turing Test, AI and Computer Security

Speaker:	Dr. Jeff YAN
		School of Computing Science
		University of Newcastle
		U.K.

Title:		"Turing Test, AI and Computer Security"

Date:		Monday, 11 October 2010

Time:		4:00pm - 5:00pm

Venue:		Lecture Theatre F (near lifts 25/26), HKUST

Abstract:

Proposed by Alan Turing in 1950, the Turing test has been a classical
method for measuring the progress of artificial intelligence research. In
the recent years, a form of automated Turing test, where the tester is a
computer rather than a human, has found widespread applications on the
Internet. Such automated Turing tests are built upon unsolved AI problems,
but used for security purposes, such as defending against spam and other
online abuses by automatically differentiating between human users and
software robots. The use of hard AI problems this way is expected to
achieve a nice win-win situation: either the problems remain unsolved and
serve as a security mechanism, or the problems are solved to advance AI
research with the collective efforts of both good and bad guys - a basic
assumption in computer security is that adversaries can be arbitrarily
smart.

In this talk, I will share some fun and challenges that we have had in
examining the design of automated Turing tests. This area was dominated by
computer vision and pattern recognition researchers before we stepped in.
Our work shows that security engineering expertise can make a unique and
significant contribution to the design of robust automated Turing tests,
as evidenced by our success of breaking a large number of CAPTCHAs,
including the schemes designed and deployed by Microsoft, Yahoo and
Google.

********************
Biography:

Jeff Yan is on the faculty of computer science at Newcastle University,
England, where he leads the Laboratory of Security Engineering and is a
founding research director for Center for Cybercrime and Computer
Security. He has a PhD from Cambridge University, and holds a visiting
post with Dept of Information Engineering, Chinese University of Hong
Kong.