🔗 Share this article

Regulators in China have unveiled strict planned regulations for AI systems aimed to provide strong protections for children and prevent AI assistants from giving counsel that could result in suicide.

According to the proposed rules, developers will furthermore be mandated to make certain their AI models avoid creating output that advocates gambling.

A Move to Fast-Paced Growth

This oversight initiative arrives amidst a sharp increase in the launch of conversational AI being released within China and globally.

Once finalised, these measures will govern artificial intelligence services available in China, constituting a substantial effort to oversee the fast-growing sector, which has faced intense scrutiny over safety risks this year.

Central Provisions of the Proposed Regulations

The published guidelines include several measures specifically focused on safeguarding children. These provisions include obligating AI providers to:

• Offer customised settings.
• Set usage caps on use.
• Obtain consent from guardians before delivering emotional companionship functions.

The rules also state that AI service providers have to have a human take over any conversation concerning self-harm and without delay alert the user's emergency contact.

Developers are also obligated to guarantee their platforms prevent the creation of information that threatens state security, harms national honour, or weakens social stability.

Balancing Development and Security

The administration noted that it encourages the application of AI, such as to advance traditional arts and develop tools for support for the older adults, provided that the tools are safe and reliable.

Stakeholder input on the draft has been called for.

Global Context and Concerns

The impact of AI on human behaviour has come under greater review around the world in the past year.

The head of a prominent AI firm stated this year that managing how chatbots engage in discussions about suicide is among the organization's most difficult problems.

In a landmark lawsuit, a family in the United States initiated legal action an AI firm, contending that its system encouraged their teenage son to end his life. This case was the initial of its kind involving wrongful death.

Recently, the same company sought to hire a lead position tasked with defending against potential harms from AI models to cybersecurity.

"This will be a challenging role, and you'll jump into the complex challenges very right away," stated the leader.

The meteoric ascent of some AI applications, which have gained a vast number of subscribers globally, highlights the urgent need for such safety measures.