China has cut America’s edge in AI benchmark scores to 3%, Bloomberg Intelligence (BI) estimates. DeepSeek’s V4.1 Flash, launched in September, pushed the gap to its lowest level on record.
That margin is the lead President Donald Trump cited last month when rejecting calls to slow AI. He claimed China is the only party pleased by the pushback against AI.
DeepSeek Moves Within 2.3 Points of Anthropic
BI put the benchmark gap at 15% earlier this year and about 9% in May. DeepSeek’s V4.1 Flash placed sixth globally on LiveBench, an independent leaderboard that scores models on questions, puzzles, and tasks.
The model scored 81.1, compared with 83.4 for Anthropic’s best entry. Still, only three of the top 15 models on LiveBench are Chinese.
The report credited deeper expertise and models tuned for domestic hardware. The gains also raise questions about whether US export curbs on Nvidia hardware are working.
Earlier this year, Anthropic argued that tighter controls could secure the US a 12-to-24-month lead by 2028. However, BI senior analyst Robert Lea cautioned that rankings shift, and Chinese labs may struggle to turn scores into revenue.
Follow us on X to get the latest news as it happens
Trump Answers AI Warnings and Data Center Backlash With China
The narrowing gap comes as the US debates how fast its own labs should advance. Anthropic CEO Dario Amodei urged pacing the frontier, a call OpenAI CEO Sam Altman and Elon Musk backed.
Data center buildouts have stirred local backlash too. An NBC News poll found 69% of adults oppose AI data centers nearby.
Trump has answered the backlash by pointing to China. He framed AI as a contest with a single winner.
“There is a SICK conspiracy going on against AI and Data Centers, and the only one that is happy about it is China. WHOEVER WINS AI, WINS!” he wrote.
He has also warned that communities blocking data centers will end up backwards and poor. Notably, Trump has turned that stance into policy.
On September 19, he announced an AI Force and vowed not to slow the industry. A White House task force under Jay Clayton, the AI czar, reports back in 120 days on AI’s risks and opportunities, along with Washington’s responsibilities.
The president signed a voluntary accord with six tech leaders that leaves safety to the companies. He has also signed an executive order to rename artificial intelligence as superintelligence. Together, these moves reflect Trump’s stated aim of keeping the US ahead of China in AI.
Subscribe to our YouTube channel to watch leaders and journalists provide expert insights
Source: BeInCrypto