At Black Hat 2016 I presented on some reverse engineering of the Philips Hue (also see my other post about getting root on it, which was part of that presentation).
If you were at the talk, you would have also seen mention that you’ll want to keep your eyes out for future publications by Eyal Ronen. You can see his website for more research related to the Hue as well, and follow him on twitter @eyalr0. He’s been doing some work in parallel that I think will do more than just R.E. the bulbs (as I did), and actually bring some of my `possible’ attacks to become real proof-of-concepts.
Summary of the work (to make it clear):
- I did NOT make a worm. The title was a question someone asked me, and the talk is about the security of the Hue.
- The mention of a possible ‘Long Range Take Over’ was new/unreleased research by Eyal Ronen – do not credit me with that. It’s part of a larger research publication that will get released at some point.
- Philips did a rather good job (all things considered). The only trade-off I really call out is reuse of encryption keys across all FW updates for all devices, which is basically what makes a theoretical worm possible.
- Rooting the Hue (earlier post) is a local attack and very nice for hardware hackers. There are unique root passwords which is a great security step, so far I haven’t found flaws in the Hue Bridge 2.0 besides that.
- There’s a lot of “interesting vectors” which the talk goes over. Given enough time some of them may give, but it’s a question of who is motivated enough to spend a lot of time on them.
You can get the full slides here too:
Philips Hue Reverse Engineering (BH Talk ‘A Lightbulb Worm?’) Slides [PDF, 8MB]
Here’s a copy of the very large whitepaper I wrote too:
Philips Hue Reverse Engineering (BH Talk ‘A Lightbulb Worm?’) Whitepaper [PDF, 5MB]
This whitepaper is a bit of a ‘data dump’ and ~48 pages of random stuff. But useful if you are interested in pushing this further!