Observations on Efforts to Quantify the Economic Effects of Counterfeit and Pirated Goods"
Here is a working URL: http://tinyurl.com/piracyreport
Jeff Sedlik, photographer, points out that without being able to quantify the amount of content piracy, which the report indicates is not possible, it is hard to estimate the economic effect. He then goes on to describe how an Image Recognition technology does seem to be able to quantify the use without attribution or permission of still images.
"I can't speak to piracy in other content arenas, but with respect to photography, advances in technology now allow image piracy rates on the internet to be quantified to an extent sufficient to estimate piracy rates with some accuracy. Image recognition technology may be used to locate instances of known images on web sites, and license data may then be used to determine whether or not each instance is authorized. Not all sites can be sampled, nor can all every instance of every image be identified, but it is possible to quantify estimated piracy rates via representative sampling.Jeff Sedlick
In 2003, PicScout http://www.picscout.com/ (an Israeli image recognition company) searched commercial web sites for instances of images of known ownership. Nine out of every ten published images were found to be used without permission or knowledge of the rights holders.
In 2005, PicScout used a new reference group of 20,000 sample images (on this occasion, provided by a group of stock photographers), and found that 1 out of every 17 copies of these images published on commercial web sites was published without the knowledge or permission of the rights holder. In the USA, the rate of misuse found in this survey was 64%. In Germany, 23%. In the UK, 13%.
PicScout reports that over a seven year period, it found that 85% of images found on commercial websites were published without the knowledge or permission of the rights holders.
In a recent LA Times article (Sept 9, 2009), Gettyimages reported that it identifies approximately 42,000 examples of copyright infringement per year, while Corbis reported the identification of approximately 70,000 infringements each year. Importantly, these figures represent only the infringements that have been detected. It is reasonable to assume that these figures represent a small fraction of actual unauthorized usages.
I am not writing to encourage or suggest heightened enforcement or penalties for piracy, nor am I expressing an opinion on copyright law, website spidering or digital rights management. I am merely pointing out that the report in question does not indicate that piracy rates are lower than estimated by industry, and that in the photography content industry, technology now allows some quantification of piracy rates. Perfect. I would not disagree with an opinion that the content industry has used piracy statistics in lobbying for support from legislators. But any attempt to claim that the figures are overstated will be frustrated by the very same issue identified in the report -- such claims cannot be quantified."
Check out the GAO report for yourself. - http://tinyurl.com/piracyreport