Fraud increased by nearly 300 per cent in the UK last year, costing public
and private sector organisations more than £900m, according to consultant KPMG.
With more than half of this figure (£468m) being stolen from organisations by
their employees, identity management and authentication is becoming ever more
critical for both consumers and businesses.
Computing spoke to Art Coviello, president and chief executive of
RSA Security, which has been tackling security threats for the past 20 years.
Q. Are criminals becoming more sophisticated in how they try to
defraud internet users and banks?
A. Phishing is still the primary way of doing fraud online,
but as technology improves we are more likely to see incidents of spyware
attacks, key-stroke logging and new types of trojan software.
Recently we have seen the invention of the hijack trojan. It waits for you to
log on so you have identified yourself, and then it takes over your banking
session and makes financial transactions from your account.
Banks tell us the level of fraud is becoming a significant problem. If they
are not already doing something, then they will soon, out of their own self
interest and because of the scale of the fraud. It is a real problem that is
only likely to get worse.
Q. Where are these internet criminals primarily operating?
A. When I go to Europe they say most of the attacks come
from the US. But in the US they say the attacks are coming from Europe or China.
The fact is that no nation or ethnic group has a monopoly on this or at the
same time are immune to it. But there are centres of excellence, and certain
countries in Eastern Europe come to mind.
The attacks are linked to the criminal entrepreneurialism of some of these
countries, but it is a problem for all of us because the internet knows no
borders.
Q. There is a lot of focus on internet crime, but surely the threat
coming from within the organisation is just as pressing?
A. Yes. In a lot of companies most people only use a static
password or static log-on, and these can be hacked in as little as 12 seconds to
12 minutes.
If you look at some of the data warehouse breaches that have occurred in the
US recently, one of them was the result of an administrator leaving a password
on a Post-It note. Internal attacks can also be a little less internal than you
think. If you have a wireless network someone can do an attack from the car park
of your building.
Q. UK banks are taking different approaches to authenticating
customers online. Do you think this could be a problem?
A. It will lead to confusion. But at the same time I don’t
think one technology should or could dominate.
There are different people and different sizes of transaction, and different
risks that banks face. It doesn’t seem possible to have a one- size-fits-all
approach for this on the internet.
Q. There has been a lot of discussion about ecommerce sites adopting
a federated identity approach. How successful has that been?
A. The technology exists today to federate identities
online, from one domain to another.
But the problem is you cannot get the back-end agreement between the banks
and the electronic retailers to pass out the liability for the identity.
In terms of privacy, we do not do anything with the consumer information, no
one needs to worry about a Big Brother approach; all the technology does is
identify passcodes.
We are the people that invented encryption on the internet, so we believe we
can protect the information we are given.
Q. You acquired fraud detection and authentication firm Cyota last
year, bringing clients such as HBOS and Barclays. What other benefits has the
acquisition generated?
A. Cyota extends our portfolio of authentication
capabilities, including a risk-based approach to authentication for banking and
online commerce transactions.
The technology looks at behaviours, origination, IP addresses and transaction
profiles. It looks for anomalies and then we adapt it based on the type of
transaction you do. If you use a different computer with a different IP address,
or you have relatives in Latvia who you want to wire some money to, for example,
it will analyse all this and decide whether or not it is genuine. The idea is
only to bother someone if an anomaly is detected.
Do you agree?
Have your say on this article