Controversial plans to build a massive database that would be capable of storing every British citizen's DNA records will be unveiled this week by the prime minister.

David Cameron will announce a groundbreaking initiative designed to position the UK at the forefront of the genetic revolution – a potentially multibillion-pound industry that is attracting attention from major technology companies, including Google.

The plan has been drawn up by the Human Genomics Strategy Group, run by Professor Sir John Bell, an adviser to the Department for Business, Innovation and Skills, with support from the Wellcome Trust, the country's largest charity.

Last week Cameron's senior policy advisor, Rohan Silva, said that the NHS's patient records were a "valuable asset" and predicted that sharing them in a safe, secure and anonymous way with researchers would allow the UK to be at the forefront of the next generation of health technologies.

"The UK has developed a life sciences strategy that is based around using the collection of NHS data … as a catalyst for life sciences innovation and growth," Silva explained in an interview. He also said that the government has learned from previous large-scale technology projects that failed.

Both the Tories and the Liberal Democrats opposed plans for a DNA database put forward under Labour, partly due to concerns that it could affect people's civil liberties.

But the coalition is expected to allow people who do not want their data stored to "opt out" – which will go some way to alleviating concerns from human rights groups.

However, there are worries about how the data will be used and shared with third parties, including commercial organisations.

The plan has been drawn up by the Human Genomics Strategy Group, run by Professor Sir John Bell, an adviser to the Department for Business, Innovation and Skills,

I think there's a problem with the general approach to how public sector bodies collect data of all types on citizens. It's a relatively new possibility, but we keep trying to crowbar it into existing state structures that have other primary responsibilities. So the police collect some and use it for policing, the NHS and the DVLA and all the others do their own thing within their silos.

It would be better to create a new structure dedicated to citizen data which would serve as repository and gatekeeper, and would grant access according to its own particular ethical standards, which take the data itself as the starting point. All agencies put their data in, nobody gets it back out without a permit. Data mining would be a special privilege and in each case must be justified, and where appropriate anonymised.

That way the benefits of having this data can be enjoyed, but the misapplication can be restricted.

We seem to have two types of complaint.
Slippery slope arguments are easily neutralised by making provision to prevent the relevant misuses, by, for instance, simply banning insurers from accessing such data on the grounds that the purpose of insurance is to mutualise risk rather than weed it out. This is easily done before or after the fact.

The principle arguments are based on the inherent lack of trust in the institutions collecting and handling the data. The problem is that lots of useful stuff can also be done with that data- especially given that a DNA database is only one example of the big data types that could become available and the real value is based on cross reference and inference. That's why I recommend the creation of a new institution, designed to be first and foremost a trustworthy custodian of this stuff.

The same can be true of civil surveillance if we put in the effort to consider what form of institution would provide the benefits with sufficient accountability. Alternatively you can swim against the tide, complaining about one database after another, while in each case 10 more that you have never heard of acquire your information and use it in secret.

And just as easily changed after the fact, should the government put it's mind to it. Once they've got the data, if they then decide to do something with it we don't like, what influence do we have to prevent it? 'We'll vote you out at some point in the future'? Fine lot of good once the proverbial cat is out of the proverbial bag. I don't want them to tell me they won't abuse it. I want the thing to be set up in a way that means they can't abuse it. Unfortunately, I can't see how that could be guaranteed.

Can you guarantee that the army will never turn their guns on parliament and install Michael Barrymore as an emperor?

If not, does that mean we must disband the army?

If not, does that mean you trust the British army?
If it was similar to the Syrian army you wouldn't trust it not to kill civilians would you?
If it was like the Egyptian army you positively expect it to overthrow the government one day.

But if, like me, you really don't expect the UK army to do those things, it might be worth spending a few moments to think about why.

After all, when the first standing army in Britain was created, those things above are exactly what everyone expected of it. And the same expectation accompanied the same institution in America. Both countries used the same solution to that. I am suggesting using it again.

I'm struggling to see the relevance of public trust in the British army as a corollary to our trust in the self-seeking nature of politicians not to sell out to the highest bidder. Are you suggesting the amount of public trust in the armed forces and politicians is, or should be, the same?

So my point is that such institutions can be created responsibly, with an internal ethic, and a duty to resist inappropriate external pressure to misbehave. Trust in politicians is not required if you have police, doctors, soldiers and so on who you do trust to refuse dodgy orders.