Stored inside a series of ordinary brick buildings beside a sprawling wasteland on the edge of San Francisco Bay are intimate details of your life, relationships and opinions.
This information repository is not the headquarters of the FBI or CIA, but Facebook Inc, Mark Zuckerberg's multibillion-dollar social networking behemoth with access to more than 840 million people, and their data.
While full-body scanners and CCTV cameras often evoke Big Brother fears, the growing trend in surveillance is much closer to home.
Social media has become the latest way governments, police and corporations spy on their citizens, most of whom have no idea they are being watched.
From the Syrian government trolling Facebook to find dissidents during the early days of the country's uprising, to commercial organisations buying users' online proclivities to sell to advertisers, social media is a snooper's honeypot.
The NSW Police scan these sites as part of routine criminal investigations, and even burglars have become savvy about social media.
''To people who think there are only innocent uses of social media, think again,'' says David Lyon, a Canadian sociology professor who has studied surveillance for 30 years.
But it is not just governments and security agencies spying on cyber space. The very nature of social media turns users into complicit, albeit low level, spies (yes, that's you, Facebook stalkers).
Historically, surveillance equipment was the specialty of security agencies and the military; everything from tracking devices to iris scanners were invented in the name of espionage.
Imagine the awe of surveillance experts in 2004 when Mark Zuckerberg and his Harvard mates launched Facebook, a program that encourages users to divulge not just personal information but their location, likes, dislikes and interests to hundreds of their ''closest'' online friends. Even the most sophisticated security agencies could not have dreamed up something like Facebook, says Lyon, the Queen's University research chair in surveillance studies, who was in Sydney this week for a conference hosted by the surveillance and everyday life research group at the University of Sydney.
''You don't even need to reveal something about yourself,'' he says. ''Anyone can go on Facebook and discover, through your friends, what your likely tastes and hobbies are because your friends often have a lot in common with you.
''It's your friends who betray you.''
After Zuckerberg launched Facebook, he capitalised on his site's mass of data almost immediately, allowing advertisers direct access to better target their products.
The information given to third parties is considered anonymous because it does not identify individuals, but the aggregation of data means companies can quickly build a detailed profile of a user.
Users can opt out of targeted advertising and request third parties delete their data. While the information people choose to post on Google+, LinkedIn, Twitter and Facebook is telling, the true value of social media surveillance hinges on friendship.
''When people go to Facebook and share a photo or a video really what is being shared is the connection,'' says inventor and social media researcher Mark Pesce.
Friendship connections reveal which people have the power to sway their friends' decisions or actions, a phenomenon known as a ''network of influence'', says Pesce, an honorary associate at the University of Sydney. ''On some topics I am an influencer, others I am influenced by,'' he says.
Friendship networks affect not only people's political affiliations and shopping habits but also their weight, smoking status and happiness. And social media has allowed these connections to move online for everyone to see, Pesce says. While he is a prolific tweeter, Pesce quit Facebook in May 2010 because the data sharing and retention policies made him uneasy. ''I didn't want Mark Zuckerberg having control of my social graph data,'' he says.
State surveillance networks have also found rich pickings online. Oppressive regimes, such as those in Syria and Egypt, have used social media sites to oust disenfranchised citizens. And the intelligence agencies of democratic nations, including the United States Defence Department and Britain's SIS, also analyse social graphs, compiled by phone and web connections, to understand terrorist movements, Pesce says.
While David Lyon acknowledges the need for police and security agencies to scrutinise criminal activity, wherever it takes place, the public nature of social media makes it simple for police to exploit these avenues of information without a warrant or court order.
During the Vancouver riots last year, police accessed images and video uploaded through Facebook and YouTube by innocent bystanders to identify rioters, a practice British police intend to imitate.
Given the role social media played during those riots and last year's Arab Spring, it is unsurprising police agencies patrol the social dialogue of the web. Last year, NSW Police made 863 requests for information from overseas social networking and internet agencies, up from 668 requests in 2010 and 383 in 2009.
''We have learnt a lot from the experience of authorities who dealt with the recent riots in the UK,'' a police spokesman said.
The agency uses Facebook, Twitter and YouTube to find witnesses, track down wanted suspects and locate missing persons, he says.
But it's not just police who benefit from online patrolling.
A British survey of 50 former criminals found most burglars case Facebook, Twitter and other sites before committing a robbery.
In the criminal's opinion, posting locations on the internet was one of the top five ways homeowners could increase their risk of a break-in.
Nash Petropoulos, who teaches media and film studies at Swinburne University of Technology, worries internet users who have grown up with social media are oblivious to the data they divulge, and happily reveal their whereabouts by ''checking in'' to venues.
''It's a generational issue,'' he says.
Foursquare, a mobile app that allows users to check themselves or friends in at various locations such as restaurants and bars and be rewarded with discounts, was one of the first mainstream location-based services when it launched in early 2010. Facebook's Places soon followed.
Now hundreds of sites make use of the GPS capabilities of smartphones to add another dimension to online socialising, including the location-aware apps for love (or lust), such as Grindr, which helps gay men meet other gay men in close proximity, and Blendr, the heterosexual equivalent.
Petropoulos believes a combination of ignorance, complicity and the trade-off between privacy and functionality makes people reveal intimate details about their life.
He says social media sites frequently change their privacy policies, and the default setting is always to make your information public. ''People need to become more privacy literate,'' he says.
However, keeping information private goes against everything that makes social networks what they are.
According to Lyon, websites such as Facebook always want more data, and members find themselves always seeking it. Users are not just victims of surveillance, they partake in it, which is why the phrase ''Facebook stalking'' has become part of the teenage lexicon.
''It's gone from stalking being a very negative word, to a joke,'' he says.
Most people, even if they choose not to admit it, have shadowed an ex, browsed the photos of their boss's kids or vetted potential housemates after reading their status rants, all in the name of online ''research''.
Lyon fears this kind of low-level surveillance will make people complacent to other types of spying.
Google+ and Facebook's use of facial recognition software to identify people in pictures appears innocent enough, he says.
But if these devices are seen as ''just a bit of fun'' they could quickly normalise other uses of facial recognition technology with more serious implications for privacy, Lyon says.
The CV you'd rather the boss didn't see
Worried those drunken 21st birthday photos your best friend uploaded online would come back to haunt you? You should be.
Peter Holland, an associate professor of human resource management and employee relations at Monash University, says employment recruiters increasingly admit to trolling social media sites such as Facebook and Google+ to fact-check details of potential employees.
A 2009 British study found 45 per cent of recruiters examined the social network sites of potential employees.
While online background checks are common practice for recruiters, employers can not discriminate against a job applicant based on their race, religion, hobbies or private activities that do not affect their job, including those less-than-flattering pictures.
''While recruiters might want to get a feel for a candidate, if you find out they have certain views of the world, you have to show you weren't influenced by that and that is a very difficult thing to prove,'' Holland says. ''You wouldn't ask someone's political affiliation [or] sexual preference in an interview.''
While there have been no cases of recruiters being prosecuted for scrutinising social media sites during their research, it is only a matter of time, says Holland, who will co-ordinate Australia's first national survey into monitoring and surveillance in the workplace next month.