Search giants must protect children from harmful content or cop $50m fine

Search engines from Saturday must have reasonable methods to determine the age of their users, and limit the risk of children being accidentally exposed to inappropriate content, or face potential fines of up to $49.5 million per breach.

The rules are part of the Australian government’s wide-ranging online safety industry codes, which passed into law last year and include a long and staggered age verification compliance timeline.

First, social media sites were required to block under-16s from December 10 last year, then from March 9 pornography sites were required to verify ages to only allow access to adults. Now, from June 27 the likes of Google and Bing search must provide appropriate experiences by determining their users’ age.

This does not mean Australians are about to be asked to record selfies or provide ID to make Google searches, since the web giant already has a good idea of its users’ ages. But it does mean providers can be penalised for failing to put adequate measures in place, so safety measures could become more visible even to adults.

“Obligations in the codes are particularly focused on preventing children’s accidental and unintentional exposure to age-inappropriate and seriously harmful material,” said a spokesperson for the eSafety Commissioner, the regulator responsible for enforcing the measures.

“Requirements are different depending on whether the user is an account holder, and has therefore had their age assessed, or if they are not logged in.”

In summary, the rules mean that search engine providers must:

• Filter out pornographic images and high-impact violence material for logged-in users who are under 18.
• Blur pornographic and violent images for logged-out users, though they can still access the content by clicking.
• Provide a referral to appropriate mental health support services as the top result for any search related to self-harm or suicide.

For logged-in users, this means adults will be able to set content rules as they like to allow or block extreme content, while children will be blocked from seeing that content. Meanwhile, logged-out users or people using the equivalent of a browser’s incognito mode will see explicit images in search results as blurred-out, unless they choose to click on them.

After several updates over recent years, this is already largely how Google Search and Bing are designed to work. Google and Microsoft have been contacted for comment.

The Age-Restricted Material Codes do not mandate the type of age assurance that services must adopt, only saying that methods must consider the technical accuracy, robustness, reliability and fairness of the solution. Options include facial age estimation and ID checks, but the tech giants are likely to rely on the AI-based age estimation that they already employ on accounts.

When creating a Google account for example, the user can provide their own age and easily lie. But as the user navigates the web, uses services, sends messages and interacts with others, metadata can help indicate a likely age range and could potentially trigger further age assurance measures.

Of course, children determined to see explicit images don’t have to lie; they can just log out or activate incognito mode. Parents would need to activate a filter on their machine, or on their network, to block the material entirely.

The next key date on the age-assurance timeline is September 9, when app stores including Apple’s and Google’s will be required to enforce age ratings, blocking underage access to R-rated software.

Get news and reviews on technology, gadgets and gaming in our Technology newsletter every Friday. Sign up here.