Agence France Presse's copyright infringement lawsuit against Google highlights how unforeseen and misunderstood technological issues can impact cases involving the Web.
Almost 18 months after AFP sued in a Washington, D.C., court, the French news agency is still struggling to collect evidence for its allegations, mostly due to the evanescent nature of Web pages. As a result, the case's discovery process has dragged on much longer than anticipated.
Meanwhile, despite its pledge to the contrary, Google seems incapable of fully ridding its Google News site of AFP material, possibly harming its defense and increasing its potential liability.
"It's really fascinating to think about the challenges that Internet subject matters create for ordinary legal procedures," said Peter Jaszi, a copyright expert and law professor at American University in Washington, D.C. who isn't involved in the case.
Beyond its potentially significant legal implications, the case illustrates how the online medium can trip up lawyers who are experts at their trade but IT amateurs.
"One of the problems with this case, and for anybody dealing with copyright in the future, is the transient nature of things on the Web and capturing the infringing conduct to prove it later on," acknowledges Joshua J. Kaufman, an attorney with Venable who represents AFP in this case.
Legally, the issue is whether Google is guilty of copyright violation by including AFP material -- such as text, photo thumbnails and headlines linked to articles in external Web sites -- in the English-language and French-language versions of the Google News search and aggregation service.
Google needs to get AFP's permission and pay AFP licensing fees to use that material, AFP claims. But Google says that it is protected by the fair use principle, which allows for limited use of copyright material. Google also argues that copyright law doesn't protect headlines, text snippets and thumbnail images.
Going into the case, AFP staffers and lawyers only printed and saved a small sample of Google News Web pages with instances of alleged infringement. They figured Google stored Google News Web pages, and that they would get access to those files during the discovery process.
They were wrong. Google kept no such files. AFP found itself without evidence of the broad infringement it had alleged. Google filed a motion to dismiss the case. "We thought Google had backed up Google News," Kaufman said. "That's why AFP didn't have interns sitting [at computers] doing 'screen print' and 'save.'"
Still, at the urging of the judge, Google and AFP jointly drafted a plan in January to collaborate on the production and identification of potential evidence. Google agreed to reconstruct 336 Google News home pages from random days between August 2003 and July 2004, and between October 2004 and March 2005. AFP then would identify instances of alleged infringement.
The joint plan quickly ran into technical problems. Google reported that for the latter part of the sampling period, it couldn't reconstruct any page. AFP and Google agreed to limit the sampling to the earlier period.
Google went back to work but another technical hurdle appeared. Google said that it had no thumbnail images stored. The best it could provide was the Web site address to which the thumbnail had been linked.
AFP took the reconstructed yet thumbnail-less pages and began what it described in a court filing as "an extremely complicated and incredibly time-consuming process."
AFP found most of the thumbnail addresses led to pages no longer online. It had to manually check if each headline and accompanying text snippet belonged to AFP. AFP enlisted its lawyers in Paris, employees in its Washington, D.C. office and outside paralegals for this task, which involved "an incredible amount of time and expense," it said.
By last week, it had identified a previously established minimum number of allegedly infringing photos, French text snippets and French headlines, but not enough English-language headlines and text. It continues to work on this.
Google also is grappling with technical problems of its own. Google pledged days after the lawsuit was filed that it would stop featuring AFP material on Google News, in accordance with its policy to comply with opt-out requests. However, a quick check by IDG News Service this week found links to recent AFP articles on Google News.
Google has declined to explain why it hasn't been able to comply with its own policy in AFP's case, saying only in a statement that "if and when we learn that there are issues with the technological filters we use to implement our opt-out policy, we endeavor to address them." The lawyer representing Google declined to comment.
The situation could expand the scope of potential infringement and puts in doubt whether Google would be able to comply with a court-ordered injunction, Kaufman said.
AFP could have taken advantage of this situation for its evidence-gathering purposes. However, Kaufman acknowledged that, until this week, he and AFP believed AFP material could no longer be found on Google News, so they weren't looking there for evidence.
For now, AFP and Google continue grappling with the technical hurdles in the case. "It's just part of learning to litigate on the Web," Kaufman said.