SYDNEY, Australia: Australia's internet regulator has warned it could move against search engines and app stores that provide access to artificial intelligence services failing to verify users' ages, as the country presses ahead with one of the world's toughest AI crackdowns.

The threat follows a Reuters review that found more than half of the most popular text-based AI services had not publicly outlined steps to comply with new rules set to take effect next week.

Australia has already taken a hard line on online harms. In December, it became the first country to ban social media for teenagers, citing mental health concerns, a move that drew praise from some world leaders who signaled they may follow suit. Now, Canberra is extending similar safeguards to AI tools.

From March 9, internet services in Australia, including search tools such as OpenAI's ChatGPT and companion chatbots, must restrict users under 18 from accessing pornography, extreme violence, self-harm, and eating disorder content. Companies that fail to comply face fines of up to A$49.5 million (US$35 million).

"eSafety will use the full range of our powers where there is non-compliance," a spokesperson for the commissioner said, including "action in respect of gatekeeper services such as search engines and app stores that provide key points of access to particular services".

AI companies globally are facing growing scrutiny and lawsuits over claims that they failed to prevent — or even encouraged — harmful behavior. OpenAI and chatbot startup Character.AI have faced wrongful death lawsuits tied to interactions with young users. OpenAI also acknowledged this week that it deactivated the ChatGPT account of a teen mass shooting suspect in Canada months before the attack, without notifying authorities.

Although Australia has not reported chatbot-linked violence or self-harm, the regulator said it has received accounts of children as young as 10 using AI chat tools for up to six hours a day.

eSafety was "concerned that AI companies are leveraging emotional manipulation, anthropomorphism and other advanced techniques to entice, entrance and entrench young people into excessive chatbot usage", the spokesperson said.

Major technology platforms may also come under pressure. Apple did not respond to a request for comment, but on its website last week said it would use "reasonable methods" to stop minors from downloading 18+ apps in Australia and other jurisdictions introducing age restrictions, without detailing those methods. Google declined to comment.

Jennifer Duxbury, head of policy at Internet Industry Group DIGI, said the regulator had been attempting to notify chatbot services of the new rules. But, she added, "ultimately any service operating in Australia is responsible for understanding its legal obligations and ensuring it meets them".

Compliance Lagging Ahead of Deadline

A week before the March 9 deadline, only nine of the 50 most popular text-based AI products had rolled out or announced plans for age assurance systems, according to the Reuters review. The assessment examined platform responses to prompts requesting restricted content, as well as published moderation policies and statements to Reuters.

Another 11 platforms had either blanket content filters or planned to block Australians entirely, measures that would comply by preventing restricted content from reaching any users. That left 30 services with no apparent public steps toward compliance.

Several major chat-based tools, including ChatGPT, Replika, and Anthropic's Claude, have begun introducing age assurance systems or broad filters. Character.AI cut off open-ended chat for users under 18.

Companion chatbot providers Candy AI, Pi, Kindroid, and Nomi said they intended to comply but did not elaborate. HammerAI said it would initially block its services in Australia to meet the requirements.

However, most companion chatbots lacked working or planned age checks, and some did not even publish an email address to report suspected breaches, another requirement under the new code.

Elon Musk's chat-based search tool, Grok, had no age verification measures or text-based content filters, Reuters found. Grok's parent company, xAI, did not respond to a request for comment. The tool is under investigation globally over concerns about the production of synthetic sexualised imagery of children.

Lisa Given, director of RMIT University's Centre for Human-AI Information Environments, said the findings were unsurprising.

"Most of these tools are being designed without a view to potential harms and the need for those kinds of safety controls," she said.

"It feels as though ... we're beta testing all of these things for these companies, and they're trying to see how far society is willing to be pushed," she added.