Slashdot Mirror


Experts Hack Power Grid in Less Than a Day

bednarz writes "Cracking a power company network and gaining access that could shut down the grid is simple, a security expert told an RSA audience, and he has done so in less than a day. Ira Winkler, a penetration-testing consultant, says he and a team of other experts took a day to set up attack tools they needed then launched their attack, which paired social engineering with corrupting browsers on a power company's desktops. By the end of a full day of the attack, they had taken over several machines at the unnamed power company, giving the team the ability to hack into the control network overseeing power production and distribution."

2 of 302 comments (clear)

  1. I'm Shocked! by ookabooka · · Score: 5, Interesting

    Not really though. A good team of social engineers (con men) and CS people can accomplish many many things...How can you prevent such things? Ridiculously strong security? Require the security guard at my place of employment to scan my ID each and every time I walk in the building? Is he supposed to also stop law enforcement from going in without clearance from HQ? I'm quite serious, what would be an effective way to stop these tactics? Everything I think of is either too impractical for most situations or prone to the same failures, but at different points.

    --
    If you are about to mod me down, keep in mind that this post was most likely sarcastic.
  2. Re:Is everything on the internet? by utunga · · Score: 5, Interesting

    I worked at a place that supposedly had two totally separate networks - one connected to the internet, one corporate wide, for news/data/intranet stuff.

    So, sure, everybody has two desktops.. one for internal one for everything else. It was great in theory - really stupid in practice. Just doesn't work.

    Reality is - there is an expectation that data from outside is available inside. In the power company case it might be everything from the latest gas pricing information to weather reports to who knows what else - and so in 'getting things done' this will inevitably require connections between the outside and the inside.

    So, as a result of this 'blanket policy' contrasting with the 'real world' people would circumvent the rule - but do it in stupid, sneaky ways -- for example in one data center there was, literally, an infrared tunnel between two computers -- "see, they are not 'physically connected' !!" .. And try to keep it secret from the network ops guys, of course.

    It would've made a lot more sense to supply a safe, heavily controlled/monitored firewall that connects outside to inside and let the network security people manage it. Otherwise your choices are (1.) actually enforece the rule and totally cripple the effectiveness of the internal system (with the result that nothing of any importance gets put there) or (2.) really lame hacks pretending to be secure and working around the blanket rule, when in actual fact they are invisible bridges that the network ops guys don't know about.

    I saw the alternative 2. in real world practice. Lets consider option 1. - if they really did manage to make the SCADA network totally seperate **and enforce that**. In that case you'd probably just end up with the forecasting/power-station-scheduling app running on the 'outside' network - and just the final 'implement it' step on the internal SCADA. Since the scheduling app is the one where the real decisions are made - hacking into that would let you send signals and information that would look relatively harmless but would still, in effect shut down the power grid. You are still sending information - in this case mediated by human brains, but not in a way that the human brain can easily understand because its low level commands (turn this up, turn that down) - that could very effectively mess up the voltage balance or frequency timing or whatever, and causing rolling blackouts and thus achieving the same aim of shutting down the power grid. There is information flowing from outside to inside - whether it is via human or machine.

    Security through dis-connectivity is a dangerous myth in most cases. In some cases, say military situations where you are willing to absorb the huge cost to re-implementing a complete replacement for just about every dang thing you might need on the inside (e.g. weather data, or radar data, say) then it may make sense. In just about every realistic corporate case - even power companies - its likely to only cause people to take their eye off the ball of implementing real security and proper firewalls etc.