A new feature in the upcoming versions of iOS and macOS, Apple’s operating systems for iPhone and PC, promises to give the “captchas” a boost once and for all. Called “automatic verification”, the technology will allow websites to verify that you are not a robot without having to do anything at all. Captcha – this is the “Fully Automated Public Turing Test to separate computers and people” – are the little tests you sometimes see when you sign up for a site to stop fraud. It may ask you to locate all the lanterns in an image or enter some oscillating letters and numbers. If you get it wrong, it may ask you to start over, leaving you wondering if you really know what a traffic light looks like – or if you end up being a robot. “You probably don’t enjoy being interrupted,” said Tommy Pauly of Apple. “It simply came to our notice then. The reason for these experiences is to prevent fraudulent activity. If you run a server, you do not want it to be flooded with scams. Some attempts to create accounts or purchase products come from legitimate users. “But other efforts may come from attackers or robots.” The company partnered with Fastly and Cloudflare, two companies that manage the infrastructure level of much of the public internet, to create the feature. It builds on the same kind of technology that underpins Apple’s efforts to override Internet passwords and works by allowing your device to send an encrypted statement confirming that it is being used by humans on the site that requests it. Although the service is connected to Apple’s iCloud network, the requesting website will not receive any personal information about the user or device. While Apple was the first to introduce such technology to the users themselves, the basic idea has been used by Google, which helped develop the template and has integrated a similar system into Chrome. However, Google’s version so far has focused on allowing third parties to create their own Captcha replacements, rather than shutting down the technology altogether. In fact, Google may even lose out on the shift: since the company bought a startup called reCAPTCHA in 2009, it has used human input from testing as part of its training data for large learning projects, asking first people to help her translate scanned books and later use the answers to train mechanical vision systems on road features in order to perfect her work with autonomous cars.
