🔗 Share this article

Authorities in China have proposed strict draft guidelines for AI aimed to establish strong measures for young users and halt AI assistants from giving advice that could result in self-harm.

According to the planned framework, developers will furthermore be mandated to ensure their systems prevent the production of output that promotes gambling.

A Response to Fast-Paced Growth

This oversight initiative arrives amidst a significant rise in the proliferation of conversational AI being introduced within China and around the world.

Once enacted, these regulations will cover AI products and services available in China, marking a substantial move to govern the rapidly expanding sector, which has come under intense scrutiny over user safety risks recently.

Key Requirements of the New Rules

The released draft rules contain several requirements particularly designed for shielding minors. These steps include obligating AI companies to:

• Supply customised preferences.
• Set time limits on use.
• Obtain consent from guardians prior to offering companionship support.

Furthermore conversational AI firms must have a human take over any dialogue involving suicide and without delay inform the individual's guardian.

Developers are also obligated to ensure their systems prevent the creation of information that endangers state security, damages national honour, or weakens social stability.

Weighing Innovation and Safety

The authorities noted that it supports the adoption of AI, including to promote cultural heritage and build tools for care for the senior citizens, provided that the tools are secure and trustworthy.

Stakeholder input on the proposals has been solicited.

International Context and Concerns

The effect of AI on human behaviour has faced increased examination globally in recent months.

The head of a leading AI firm stated this year that managing how AI systems deal with conversations related to mental health crises is among the organization's most difficult challenges.

In a notable case, a family in the United States filed a lawsuit an AI developer, alleging that its AI assistant influenced their teenage son to end his life. This legal action was the first of its kind involving liability.

This month, the same company sought to hire a key position responsible for managing risks from AI models to cybersecurity.

"The will be a demanding role, and you'll jump into the deep end pretty much from the start," commented the leader.

The swift popularity of certain AI services, which have attracted millions of followers worldwide, highlights the pressing need for such governance measures.