China is reportedly set to establish wide-ranging standards for artificial intelligence (AI) safety.
That’s according to a report Tuesday (Nov. 3) from Bloomberg News, which said these measures would be among the most thorough guidelines to protect against a loss of control as AI becomes more powerful.
The report cites comments from China’s economic planner saying its ramping up legislation to govern how AI is developed and applied from development to deployment.
The regulations will extend to things like data to computing power, said Wang Ruomeng, a senior official with the National Development and Reform Commission. Bloomberg notes that this means companies like DeepSeek, Moonshot and Alibaba would be covered as they try to compete against U.S. rivals OpenAI and Anthropic.
“We will accelerate legislation to support core AI enablers — data, computing power and algorithms — and regulate key application scenarios, while establishing safety standards and ethical guidelines covering the full life cycle of AI,” Wang said at a briefing in Beijing.
He added that the government will conduct regular reviews of how the rules are carried out and revise guidelines that become outdated.
China will also keep an eye on AI’s impact on employment and displacement, Wang said, and provide protections for affected workers. At the same time, the government will keep encouraging AI development, though he called on countries to cooperate on protections around the technology.
“AI safety is not a problem one company, industry or country can tackle alone — it is a global challenge requiring international cooperation,” Wang said.
In other artificial intelligence news, recent PYMNTS Intelligence research shows that many eCommerce merchants (37%) are aware when a visit to their sites came from an AI platform, but end up losing the trail prior to purchase.
“The opportunity extends across product discovery, marketing and checkout,” PYMNTS wrote recently. “A merchant might improve product information for AI assistants, work with an eCommerce provider or participate in an agent platform. Following the resulting purchases would give each investment a measurable commercial outcome.”
The research showed that 51% of merchants identified some AI-originated visits but reported inconsistent attribution. Another 12% could not determine which traffic came from AI agents and which from human shoppers.
“Better measurement could support more specific spending decisions,” the report added. “Retailers could compare whether changes to product listings bring visitors who purchase, whether particular referral sources deliver completed orders and where shoppers leave before checkout.”
