Could the 'right to be forgotten' become automated? Oblivion software could help Google remove hundreds of people from the web in seconds By Victoria Woollaston for MailOnline Published: 15:09 GMT, 26 June 2015 | Updated: 15:10 GMT, 26 June 2015 In the year since the European Court of Justice ruled that anyone can ask Google to remove personal information about them, the site has evaluated more than one million links. Each request has to be verified and processed by a dedicated team of people, but the sheer volume can cause delays.To speed this up, researchers from Germany and New Zealand have developed an algorithm capable of analysing hundreds of such requests in seconds. And they hope to offer it to Google, and other search engines, to help them manage future demands. Called Oblivion, the software allows a user to automatically find and tag their personal information on the web, using both text - or natural language processing (NLP) - and image recognition.They are then taken to a form where they can securely and privately file their request with the relevant URLs. To confirm their identity, the user sends a digital copy of their ID to what the researchers described as a certificate authority (CA) - a separate body designed to securely handle the data. This software confirms their name, age, address and nationality. Oblivion will then check the webpages the user wants to remove and tag relevant references.Following the tagging process, Oblivion sends the request and tagged pages to Google who confirms the details match with those in the article, and if they do, the user is given an 'ownership token'. These tokens are then linked with the information on the request form and both are submitted to Google's takedown team. The researchers, led by Milivoj Simeonovski from Saarland University in Germany, have tested Oblivion on existing articles and found it to be capable of handling 278 removal requests per second. This is on a standard notebook running a 2.5 GHz dual core processor, suggesting more powerful computers could process higher numbers.The final takedown decision still needs to be made by a human because the variables, including the public interest defense, needs to be handled on a case-by-case basis.But Oblivion would speed up this process and confirm the validity of the claims within seconds. 'Search results typically comprise a variety of sources that contain personal information - either intentionally released by the person herself, or unintentionally leaked or published by third parties without being noticed, often with detrimental effects on the individual's privacy,' explained the researchers in their paper Oblivion: Mitigating privacy leaks by controlling the discoverability of online information. 'To grant individuals the ability to regain control over their disseminated personal information, the European Court of Justice recently ruled that EU citizens have a right to be forgotten.At the time of writing, Google has evaluated 1,002,790 URLs submitted since 29 May 2014 and removed 41% of those it has fully processed (interactive module above). The final takedown decision still needs to be made by a human because the variables, including public interest claims, needs to be handled on a case-by-case basis'As of now, these technical means consist of a web form that requires a user to manually identify all relevant links herself upfront and to insert them into the web form, followed by a manual evaluation by employees of the indexing system to assess if the request to remove those links is eligible and lawful.'In this work, we propose a universal framework Oblivion to support the automation of the right to be forgotten in a scalable, provable and privacy-preserving manner.'At the time of writing, Google has evaluated 1,002,790 URLs submitted since 29 May 2014 and removed more than 41 per cent of those it has fully processed.Last week France's data watchdog gave Google just two weeks to extend the 'right to be forgotten' ruling to all of its domains.The agency CNIL issued the order because Google only deletes the information from searches made through European domains, such as google.fr or google.es.By switching to a non-European Google domain, such as Google.com, internet users can still access the deleted links relatively easily. French citizens have made more requests than from any other European member state, with 55,000 requests.Google added that Facebook accounted for the largest number of deleted URLs. But not all requests are granted, only those deemed by Google to be valid. Isabelle Falque-Pierrotin, president of CNIL, said the order to remove search results globally 'is only telling international companies that operate in Europe that they must conform to European law'.Google said that it is working to comply with the EU decision, but did not say whether it would apply the French order.'We've been working hard to strike the right balance in implementing the European Court's ruling, co-operating closely with data protection authorities,' the company said in a statement.'The ruling focussed on services directed to European users, and that's the approach we are taking in complying with it.' If Google does not meet the deadline, it faces a relatively insignificant maximum fine of €150,000 (£110,000).However, it will also be forced to put a banner on its homepage saying it had failed to follow French law.The CNIL added that if Google fails to meet the deadline, it would begin compiling a report about the company that would be used to decide if the search engine should be sanctioned.