DARPA calls Big Data boffins: Help us lock up everyone's privates The American military is looking for number-crunching wizards able to tackle the national security threat posed by, erm... publicly available data. The Defense Advanced Research Projects Agency (DARPA) is searching for boffins to "measure the national security impact of public data and to defend against the malicious use of …
Make it so that people can have anonymous vested identities. Make it impossible through a combination of legislation and technical means for anyone to assemble a database of personal information that the subjects have not (with *real informed consent*) opted *into*.
Set expiry times by default on personal data so that unless there is an ongoing relationship and another opt-in, the data melts away.
Make it impossible for third parties to access or tamper with data:
Example:
Party A Creates Customer Name/VotingKey list
Customer gets VotingKey
Party B gets list of VotingKeys
Party B Creates Ballot/VotingKey list
Voter uses VotingKey to obtain and register Ballot.
Party A knows the customer had the right to vote, but not what they voted.
Party B knows what the customer voted, but not who they are.
Party B publishes votes and keys so that voter can check that their vote is counted as cast. Public can verify there are no more ballots than their should be.
The system is open to third-party audits without revealing how people voted.
To the extent that there is a vulnerability to collusion you just use the same principle to extend to more parties, including the entire population if it comes to it. If everybody is required to release the information
There are a wide variety of ways to partition data so that only the legitimate subject has control. The problem is twofold:
Most, including politicians and even technical people do not understand this stuff. More people who can understand need to skill up.
Those who *do* understand have vested interests in the status quo.
The technical ability is there in spades. This is entirely a matter of political will.
The problem is that the customer gets the voting key and, later, "Party B publishes votes and keys so that voter can check that their vote is counted as cast" That means it is no longer a secret and anonymous ballot because the voter can prove which way they voted. That makes voters open to pressure and allows them to sell their votes. A genuinely free vote depends on people NOT being able to prove how they voted.
I'm not so sure of that. As the article says, the NSA already has a handle on how to use big data, seeing as the PRISM surveillance programme managed to collect the details of millions of people every day, but that isn't the same as being able to do anything with it.
It strikes me that the data boffins, in doing what DARPA wants, will devise ways to maximise the usefulness of the data they play with, and I'm sure the NSA could feed off that.
Took a look at the referenced document.
We know that the NSA has assembled massive amounts of data and they even admit to having all the metatdata.
Big Brother has all the raw data.
Big Brother knows how to decrypt things using Rubber-hose cryptanalysis.
Big Brother only needs a little help in the subtleties of how to mine their data.
They *say* they are assessing vulnerabilities in large bodies of data, but they want awfully specific functioning tools to attack the data themselves as a proof of vulnerability.
From the referenced document:
"The goal of this topic is to develop tools to characterize and assess the nature, persistence, and quality of data. ... Also of interest are tools to characterize the quality of data for automated processing and analysis (i.e., a measure of how much manpower would be required to use a specific source). ... Additionally, the goal of this topic is to characterize the threat through the creation of tools, techniques, and methodologies to measure the vulnerabilities in a given set of public data. As an example, reconstructing the profile of an organization from many data pieces using low computational-complexity methods might indicate vulnerability. "
This actually makes a lot of sense.
One of the purported interesting use cases of big data it to do exactly what is being talked about here, to draw inferences, to make meaningful generalisations, for distinct ideas to emerge out of the sea of data.
That someone could learn interesting things about a government or a nation from the data it outputs (e.g. census, taxation, infrastructure, energy, financial etc) is one of the logical consequences of the data being available.
That someone with wicked intent might find out something useful from it is not just *possible*, it is pretty inevitable. *How* useful, well ... dunno.
When the Gestapo come to your house and ask you to denounce the "traitors" that live in your neighbourhood you have a choice.
Squeal or die.
Choose die and you will not regret it. Even if they threaten to kill your children in front of you.
For this is the ultimate test of faith.
Satan can kill your body but only God has the power to destroy your soul..
Leaving imaginary beings out of it, I sort of agree with you. The trouble is, I doubt the (putatative) powers of totalitarianism will be so crude as to mere kill these days. There are ways of hurting that don't kill, and leave all involved ruined. I won't go into detail - I doubt I need to.
Finally, DARPA wants to design a real-world tool that can monitor open source data sets in real time, measure vulnerabilities andthen provide defensive countermeasuresexploit those vulnerabilities to supplement the data acquisition capabilities of the NSA through its PRISM and XKeyScore programs.
There...fixed it for ya.
Right now researchers are working on ways to anonymize data. DARPA wants to be sure they can de-anonymize the data, and what better way than to have an inside line on how it will be anonymized in the first place.
Also plenty of big data has been available for years, but until now, was not seen as a danger. Phone books, mailing lists, census data, auto and license lists, and on and on. It isn't the data that has them worried, it is the anonymizing.
Winston and his colleagues can be employed usefully re-writing all Public Big Data to ensure it can only be used in ways useful to Big Br.... er The US Government if all dat is vetted and suitably amended it will be safe!
The Ministry of Truth has my vote!
I see all these posts about how DARPA is going to use any of this new tech to better get the squeeze on our privacy, and I understand very well where that is coming from and I blame nobody.
But let's take a step back just one second.
Look at cryptography. I've always heard that a secure cryptographic system is a system where everyone knows exactly how it works, because knowing how it works does not mean you can break it.
So, if we apply that rule to data anonymisation, I think we could very well design something where data is effectively anonymous and could not be recombined to something that would break anonymity.
I don't think it will be easy though.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
