AusCERT 2011: Second Stuxnet worm on the horizon

Eric Byres says the worm has created “an arms race”

The rogue Stuxnet worm that infected an Iranian nuclear facility two years ago has provided a blueprint for cyber criminals and we can expect to see another one in the future.

That’s the view of US-based Byres Security consultant, Eric Byres, who told delegates at AusCERT 2011 that it was highly likely that we could see the Son of Stuxnet soon.

The original worm managed to infect machinery in an Iranian nuclear plant in June 2009, with the infection vector believed to be flash memory. It took the initial variant, compiled on June 22, 2009, just 12 hours to infect its first PC.

The worm infected at least 100,000 computers and at least 22 manufacturing sites.

“We have created an arms race because now countries like China are blaming the US for the Iran attack and saying we need one too,” said Byres. “I think the next Stuxnet will be cruder but it will go after broad spectrum connections.”

According to Byres, the original Stuxnet may not have got into the nuclear plant by way of USB stick.

To prove this, he began working with firewall manufacturer ,Siemens, and released the worm into the vendor’s supervisory control and data acquisition (SCADA) system.

“The first thing was that there were very few initial starting points,” said Byres. “People assumed it was the USB key that got to that host.”

While the USB stick was very useful, he said it was likely the contractor could have given the client infected project files.

“It was almost a sure bet that would go through,” he said. “It could have been transmitted as a drop package.”

To prove this, Byres showed delegates a CD-ROM disc which contained the Stuxnet worm. When the company let the worm infect live systems in the Siemens lab, there were checks done. “Once the worm got out, he had many pathways to get in,” he said. “The worm started analysing things we didn’t know about like shared network drives and print spooler services. It was able to exploit those drives and infect other computers."

Tracking the worm proved complex as the number of pathways it exploited started to "explode" and by the end of the week Byres had discovered 17 pathways of exploitation.

“People say we need to 'airgap' systems [cut them off from the Internet] but there are patches and anti virus signatures that need to come in. Even the best airgap is an illusion.”

Byres also warned there are no patches available for Stuxnet because it was a design, not a bug, and the security industry needed to look at safety functionality.

“Rather than control the whole system, we should protect mission critical equipment. Combining safety functionality with controls is making it easier for Stuxent,” he said.

“The worm is acting like a training ground for cyber criminals. It’s showing them what can be done, and that’s why I believe son of Stuxnet is not far away.”

His comments echo those of Lofty Peach consultant, Ron Southworth, who told delegates at AusCERT 2011 that Australia needed to build a universal security system to prevent attacks from worms such as Stuxnet.

Hamish Barwick travelled to AusCERT 2011 as a guest of AusCERT

Got a security tip-off? Contact Hamish Barwick at hamish_barwick at idg.com.au

Follow Hamish Barwick on Twitter: @HamishBarwick

Follow Computerworld Australia on Twitter: @ComputerworldAU

Tags siemensStuxent wormauscert 2011Eric Byres

More about CERT AustraliaetworkPeachSiemens

2 Comments

Frank

1

Protection from attacks such as Stuxnet are already available with many organisations already protected when this broke out. Like with the Sony attack, there are security solutions already built, you just need to access them on an organisational level and not just a point product ,"defence in depth" approach of the early 2000's.

Eric Byres

2

Sorry, but I have to disagree - protection from Stuxnet might be available now, but I am puzzled as to what protection would have been available prior July 2010 for a real world SCADA or ICS deployment.

And I don't think there is protection from the next Stuxnet because we don't know what it will look like. For example, what would you suggest to protect a Siemens S7 system from the vulnerabilities NSS labs has been discussing over the past week? I just don't see a simple solution here. I wish it was different, but after 30 years in designing SCADA systems, I see a big problem in front of us.

Comments are now closed

Queensland Police arrest man for allegedly hacking US gaming developer site

READ THIS ARTICLE
DO NOT SHOW THIS BOX AGAIN [ x ]