Retailers could guess your age, sex, birthday and wedding anniversary simply from the types of gifts purchased for you online and their timing, according to a patent granted to online retail giant, Amazon.
The information could be used to remind your loved ones of an impending special occasion and offer gift suggestions.
Currently Amazon makes personalised suggestions to customers based on previous purchases by that customer, previous web pages browsed and comparisons between customers who have bought similar products. But the company may vastly increase its predictive capability in the future.
The patent describes software that automatically guesses when a gift is being purchased by extracting key words such as "birthday" or "anniversary" from an attached message. It might also note details such as the fact that the buyer has asked it to be gift wrapped or that the recipient address is different from the purchaser address, according to the patent, which was granted on 8 March.
The software would then infer the recipient's age and gender according to the type of gift, the paper it is wrapped in and by cross-referencing any past appraisals of the items purchased. Amazon would remind potential gift purchasers by sending them emails or an alert when they log on to the website.
"Customers may be provided with reminders of important events, thereby avoiding the embarrassment that ensues when one forgets such events," the patent declares.
But privacy advocates are outraged by the idea because it would profile people who may never have used Amazon, says Jason Catlett, whose website Junkbusters, based in Green Brook, New Jersey, US, raises consumer awareness of marketing activities. Furthermore, he says: "All databases of information become honey pots for identity thieves."
Most worrying is that the patent appears to target children, says Karen Coyle of the public interest alliance Computer Professionals for Social Responsibility in Berkeley, California.
According to Amazon spokesperson Patty Smith, based in Seattle, Washington, these worries are "a little premature and a bit speculative". She adds that the company has no plans to implement the technology at present.
But making judgments about customers based on information harvested through existing interactions is part of a broader trend in retail, says Mark DeSantis of Hyperactive Technologies in Pittsburgh, Pennsylvania.
"We glean whatever clues we can from the information that is coming in," he explains. "Behaviours seem complicated on the surface but you can actually make pretty accurate judgments about people. I wouldn't be surprised if Amazon gets really good at this."
Hyperactive's software, called Hyperactive Bob, is currently slashing waiting times and wastage in US fast food restaurants by predicting with 90% accuracy what food customers will order from the type of car they drive to the store in.