Tweet
A new study shows that 28% of piracy takedown requests submitted to Google may be questionable at best.
The study by researches at Columbia University's American Assembly and Berkeley's School of Law looked at more than 108 million takedown requests submitted to the search engine, and found that a large percentage of these may be invalid.
4.2% of these requests targeted content that wasn't even present on the listed URLs, while another 24% raised issues about the request's validity, including in relation to fair use.
Some also targeted websites that no longer existed, such as the now shuttered Megaupload.com.
Rights-holders themselves usually aren't involved in the actual process of locating infringing links. Instead, they outsource the job to companies that specialize in anti-piracy activities. These companies will normally use automated scripts to scour the web for potentially infringing links, but these automated "bots" are prone to false positives.
Worse yet, when these links are submitted to Google, Google's uses their own bots to determine the validity of the request, meaning the entire process could be devoid of human intervention and verification. The researchers also say Google, like many other service providers, takes a cautions approach when it comes to removal requests and often over removes content just to protect themselves legally.
This, the researchers say, is why calls for a "take down, stay down" regime are potentially very dangerous, with many legitimate sites likely to be wrongly targeted.
The study by researches at Columbia University's American Assembly and Berkeley's School of Law looked at more than 108 million takedown requests submitted to the search engine, and found that a large percentage of these may be invalid.
4.2% of these requests targeted content that wasn't even present on the listed URLs, while another 24% raised issues about the request's validity, including in relation to fair use.
Some also targeted websites that no longer existed, such as the now shuttered Megaupload.com.
Rights-holders themselves usually aren't involved in the actual process of locating infringing links. Instead, they outsource the job to companies that specialize in anti-piracy activities. These companies will normally use automated scripts to scour the web for potentially infringing links, but these automated "bots" are prone to false positives.
Worse yet, when these links are submitted to Google, Google's uses their own bots to determine the validity of the request, meaning the entire process could be devoid of human intervention and verification. The researchers also say Google, like many other service providers, takes a cautions approach when it comes to removal requests and often over removes content just to protect themselves legally.
This, the researchers say, is why calls for a "take down, stay down" regime are potentially very dangerous, with many legitimate sites likely to be wrongly targeted.