Techno File: Google in clear

Published on Sunday 26 July 2009 22:15

GOOGLE is not the publisher of defamatory words that appear in its search results, the High Court has ruled. Even when Google had been told its results contained libellous words, it was not liable as a publisher, said Mr Justice Eady.

The search giant's US and UK operations were sued in England by a London-based training business over comments about its distance learning courses that appeared in the forum of a US website. The comments were said to be defamatory and an excerpt from them could be found in Google search results.

Metropolitan International Schools (MIS) runs distance learning courses in games development under the name "Train2Game".

Google argued it had no responsibility for the words complained of, was not responsible for the piece of text about MIS and had no liability with regard to any publication prior to notification as to the identity of specific URLs that contained the words. It said it would not be liable even after if it was notified of the URLs.

The judge concluded Google was a mere facilitator. A earlier case, the Bunt case, also heard by Mr Justice Eady, confirmed that mere facilitators, like telephone carriers, were generally not liable for defamatory content.

Highland warning

THE Office of the Information Commissioner (ICO) has taken action against a Scottish local authority which lost two laptop computers, despite the fact they were stored in a locked office and password-protected. The ICO found Highland Council in breach of the Data Protection Act (DPA). The laptops, which were not encrypted, had personal details of 1,400 people and medical information about some. The ICO said "no additional physical security measures were in place". A formal undertaking signed by council chief executive Alistair Dodds commits the council to encrypting all mobile devices containing personal data by the end of September.

Meanwhile, the ICO is to get new powers to issue fines next April. Under the DPA, it cannot issue fines for breaches of the eight data protection principles at the heart of the law. From April, that will change and it will be able to issue fines for knowing or reckless breaches.

Facebook rapped

FACEBOOK does not protect personal information well enough to comply with Canadian data protection law, the Canadian Privacy Commissioner has said.

Jennifer Stoddart's office has investigated the social networking website's use of personal information and has found Facebook is not clear enough about how users can control their information or restrictive enough in limiting other companies' access to it. The commissioner's office said that the company needed to be more transparent.

The investigation found that users were told on Facebook how to de-activate accounts, but not how to delete them. Only deleting accounts actually removes personal information from Facebook's servers.

The report also said that Facebook should not keep for ever the data contained in accounts which are de-activated – and that the consent gained by Facebook was so broad and general so as to be not adequate to absolve it of responsibility. "In my view, consent obtained on such a basis is meaningless," said Ms Stoddart's report.

Patent system block

A PIONEERING patent system being piloted in the United States has been stopped from accepting new submissions.

The Peer-to-Patent scheme harnessed web users' knowledge to improve patent quality, but the project has been curtailed. It will no longer assess new patents but will process those already on its books, which is expected to take until October.

The system was launched two years ago as an attempt to harness the wisdom of web users to ensure nobody was granted a patent monopoly on things already invented.

Patents are assessed by employees of patent offices, but one of the problems they face, particularly with highly technical patent applications, is knowing whether or not a claimed invention is truly new.

The Peer-to-Patent project aimed to solve that by allowing third parties to tell hearing officers anything they knew of that pre-dated an application and covered the same ground. This is known as "prior art". It was an experiment that depended on amateur experts submitting knowledge to patent examiners while applications were considered. This was against US patent law, which bars third-party submissions. An exception was made for the programme, but only for patent applicants who opted in to the system.