So while it's huge Google was told what to do, it's not surprising as this is business as usual. And back to an earlier point... The best place to siphon data in 2019? Your phone. Times have changed, data collection by governments hasn't.
So while it's huge Google was told what to do, it's not surprising as this is business as usual. And back to an earlier point... The best place to siphon data in 2019? Your phone. Times have changed, data collection by governments hasn't.
But can't you reverse engineer your phone and see what it's doing? And can't you monitor the network data it's sending? With a backdoor in long haul transport gear, academics, researchers, random hackers, watchdog groups, journalists, competitors, etc, don't have the ability to monitor for bad behavior.
I honestly believe that the theoretical DMA backdoor attack (and most other similar attacks) have been mitigated thoroughly. I am much more concerned about secretly held 0days (RCE) and most concerned about warrantless orders against cloud storage.
https://www.apple.com/business/site/docs/iOS_Security_Guide.... p41
Are you citing "To protect the device from vulnerabilities in network processor firmware, network interfaces including Wi-Fi and baseband have limited access to application processor memory. When USB or SDIO is used to interface with the network processor, the network processor can’t initiate Direct Memory Access (DMA) transactions to the application processor. When PCIe is used, each network processor is on its own isolated PCIe bus. An IOMMU on each PCIe bus limits the network processor’s DMA access to pages of memory containing its network packets or control structures."? Correct?
The attention to hardware isolation and separation is appreciated, but I don't hold my breath for iBoot and SEPOS protecting an iPhone from powerful adversaries.
I think that these mechanisms completely frustrate "bulk" in-field collection efforts; for example, scanning all phones at DUI checkpoints.
No technical control is perfect. If you personally piss off a nation state adversary, they are more likely to yeet you off to a black site and hit you with a wrench until you cough up your passcode.
Surely, someone will break iBoot, and surely, someone will break SEPOS. And surely, someone will chain a kernel exploit with a userspace exploit [0]. And surely, someone will leak the signing keys for a widely deployed cheap android phone [1]. And surely, someone will push 777 permissions to a cloud provider [2]. And most surely, powerful government adversaries will hold brutal exploits close to their chest in the service of power and politics [3].
So I guess, if you want to breath freely: host your infrastructure yourself where feasible. Choose providers who respect your privacy. Make a modest but financially fair donation to the EFF. Become politically active. Use better practices - not best - to avoid fatiguing yourself in the windmill chasing effort of being Perfectly Secure. Most importantly, stay awake and aware and ready to fight.
0 https://github.com/Cryptogenic/Exploit-Writeups/blob/master/...
1 https://www.theregister.co.uk/2017/11/16/dji_private_keys_le...
2 https://www.cnbc.com/2019/05/17/salesforce-says-a-major-issu...
3 https://www.wired.com/story/eternalblue-leaked-nsa-spy-tool-...
[0] https://www.cisco.com/c/en/us/td/docs/routers/10000/10008/fe...
> The surveillance is performed through the use of wiretaps on traditional telecommunications and Internet services in voice, data, and multiservice networks. The LEA delivers a request for a wiretap to the target's service provider, who is responsible for intercepting data communication to and from the individual. The service provider uses the target's IP address or session to determine which of its edge routers handles the target's traffic (data communication). The service provider then intercepts the target's traffic as it passes through the router, and sends a copy of the intercepted traffic to the LEA without the target's knowledge.
Responding to lawful warrants and subpoenas is everyone’s obligation, and has been for hundreds of years under American law, and English law before that. The government is entitled to almost any evidence—it just has to follow the proper process to get it. Lawful intercept just supports that process. That’s also why the FBI having access to US data is fundamentally different than Chinese back doors in US networks. US law enforcement has legal ways to access data flowing in networks to perform their legitimate law enforcement functions. The Chinese have no legitimate reason to access data in US networks.
First of all PRISM was not purely "a system for handling paperwork for lawful intercepts", see [0].
Second your assertion assumes that somehow all requests within these programs abided by "lawful purpose", which we also know is not true. Whether FISA rubber stamps were lawful is subjective and your opinion but not fact. Your answer purports a black and white perspective on the past which I simply don't find appropriately represents all the shades of grey presented by all of the legal angles during this time.
[0] https://dzone.com/articles/nsa-prism-%E2%80%93-mother-all-bi...
[0] https://www.theregister.co.uk/2016/04/20/public_advocate_fbi...
[1] https://www.aclu.org/blog/national-security/privacy-and-surv...
Just because it was never settled that way doesn't change the history of it.
You could have made the argument you were trying to make colorably and defensibly. All you had to do was not try to sound like you'd been "read in" to NSA's SIGINT programs. But, like I said above: you managed to cite one of the few NSA program examples that is in fact totally banal and, ironically, an almost perfect example of lawful intercept.
As I said above: it's not hard to come up with cases where NSA is doing things that appear to contravene US law, but you managed to cite the one instance where all they're managing is paperwork.
I can't imagine much useful material comes from wiretapping these days. Maybe once in a while, but the real value largely exists in the application layer, which is obtained in a different way.
Foreign-hosted services seem like they'd be hard to crack, but it's extremely likely their data flows though Cloudflare, Amazon, GCE, or a similar US-based company.
Law enforcement still had to go to a judge, get a court order, and take it to the ISP. The ISP could then configure the devices to siphon off traffic from a single customer to a collection device.
The same functionality and process has existed in phone networks basically since they became electronically switched.
Remember the context of my conversation was 2010-2011 which was pre-Snowden. It's likely state, local and national agencies have less of an interest, today, at route/switch infrastructure simply because of the post-Snowden crypto push.
[0] https://www.vice.com/en_us/article/mvp8ga/the-fbi-wants-to-w...
It was later discovered the NSA capability was the direct result of tapping fiber optic cables between international data centers, nothing to do with lawful intercept capabilities.
Tell me if I am mistaken, but in these specific case it looks more to me like forbidding customers and companies from using foreign made secure vaults because the government can't force them to secretly provide a master key.
Then sure, the argument is also valid that these vault makers could be providing a master key to their respective governments, but that would be a different argument and different enforcement.
Yes, and these actions are the US finally shooting back after enduring years of industrial espionage, forced technology transfers, and market access restrictions.
The assertion that there's going to be consequences seems... inevitable? To me at least.
After 9-11, the FBI's mandate was expanded from investigating terrorist attacks to include preventing them from occurring.