Both.news
Donate
Story AnalysisTechNews report

China says it will curb tech exports, including AI, citing national security risks

The proposal came as China and the U.S. made progress in dialogues over AI safety and governance and on the setting up of a notification mechanism between the two superpowers for AI incidents that could threaten national security.

PBS NewsHourPublished Analyzed on October 16, 2026

This page is Both News's AI analysis of the article, not the article itself. Read the full story on the publisher's site.

Our verdict

Political lean

Center
◀ Left 10Center10 Right ▶

0 = center. The further from 0, the stronger the lean (max 10).

Reporting balance

59
Somewhat one-sided
59 / 100

How complete and even-handed the reporting is — separate from which side it leans.

This is a plain, attributed AP wire report with neutral language and a lean near zero. Its main weakness is one-sidedness: it carries the Chinese ministry's rationale and two commentators, but no US, Western or affected-company response, and the headline gives national security more weight than the body does. Readers should keep in mind that the measure is a draft with no stated timeline, and should look for US and industry reactions before drawing conclusions about its effects.

Why this lean
The story has no clear left-right axis. It is a wire report on a Chinese trade-policy proposal. Its attributions are mostly neutral: the Chinese ministry's justification is given as a direct quote, and the analysts are identified by employer and attributed with hedges ('he believes', 'can be seen as'). The framing terms ('technology race', 'superpowers') are mild and do not take a side. The only tilt is in what is included: the Chinese government's stated rationale and one interpretation of it appear, while no Western or corporate response is given. That affects completeness, not political lean.
Framing
The story is told from the Chinese government's announced position, with analyst commentary on AI competition and geopolitical fragmentation, and treats national security as the central question.

Confidence: Medium — The text is a single wire story and appears complete. My assessment depends on the article alone. I cannot verify the draft's details, the parameter counts, or the Anthropic background, and I have not checked any outside coverage of the proposal.

Key Findings

  1. 1The Chinese government's justification is quoted directly and at length ('necessary', 'security bottom line', 'human control'), and the story lets the ministry state its rationale in its own words. This is a strength in attribution, although the article does not test those claims against any outside view.
  2. 2The headline says China is citing 'national security risks', but the body says the ministry 'cited national security as one of the reasons'. The headline drops the qualifier, so it presents security as the main reason when the text reports it as one of several.
  3. 3Two analysts are quoted, and both are identified by employer. The Omdia analyst's interpretation is hedged ('can be seen as', 'he believes'), which is good practice. Both analysts are commentators, though, and no official or company from the other side of the dispute is quoted.
  4. 4The counts are specific and add up (23 AI + 12 solar + 5 biomedicine = 40). This gives readers a checkable breakdown, but no baseline is given: the article does not say how many technologies were already restricted or how this changes the catalog.
  5. 5The article says the draft 'did not name specific models', yet it states that Kimi K3 and Qwen 3.8-Max have parameter counts above 2 trillion. The claim is presented as AP's own statement, with no source given for the parameter figures.
  6. 6The US context is limited to a background paragraph and a single sentence about Anthropic. The article notes that the restrictions 'were later lifted' but does not explain why. A reader cannot tell from the text how the US policy compares with the Chinese proposal.

Who gets a voice

  • Ministry of Commerce spokesperson (unnamed)
    Chinese government body that jointly published the draft and explained its rationale
    Chinese government, national security and AI safetyMain voice
  • Lian Jye Su
    Chief analyst at technology research firm Omdia
    Technology-industry analysis of Western sovereignty measuresMinor voice
  • Gary Ng
    Senior economist at French bank Natixis
    Economic commentary on tech fragmentationMinor voice

Loaded language

“technology race”→ technology competition

'Race' suggests a contest with winners and losers, which frames the dispute as competitive rather than as a policy matter.

“superpowers”→ major powers or countries

The label is widely used but carries a framing that casts the US and China as the central powers of a rivalry.

“intensifies”→ increases or grows

Suggests escalation, which the article does not document with a specific event.

Dimension Breakdown

Source Diversity5/10

Three attributed voices: a ministry statement, an Omdia analyst and a Natixis economist. All are either official or industry-commentary sources. No US, Western, company or independent-policy voice appears, so the range is narrow.

Language Neutrality8/10

Language is plain and attributed. Claims are tied to the ministries or to named analysts. Framing words such as 'technology race' and 'superpowers' are mild, and no pejorative labels for China or the US appear.

Data Balance7/10

Counts are specific and internally consistent (40 = 23 AI + 12 solar + 5 biomedicine). Attribution is given for the ministries. The article gives no baseline for the existing catalog, and it presents parameter-count claims about named models without a source.

Perspective Completeness5/10

The Chinese government's side is heard directly, and the analysts add some outside interpretation. US, Western and affected-company perspectives are absent, even though the article itself notes US-led restrictions and the Anthropic case.

Headline Accuracy7/10

The headline matches the core news of the proposal. It drops the body's qualifier that national security was 'one of the reasons', which makes the stated rationale look more singular than the text reports.

Context Sufficiency6/10

Some background is given: US-led restrictions, the blocked chips, and the Anthropic reference. The Anthropic item is one sentence with no reason for the reversal. The article does not explain what the draft's legal status means, when it could take effect, or how it compares with existing controls.

Missing Perspectives

  • US or Western government reaction: Without a response from US Commerce or trade officials, readers cannot see how Washington views the proposal or whether it intends to respond. This matters because the article itself frames the move as a response to Western controls.
  • Affected companies and industry groups: Companies such as Alibaba and Moonshot, which the article names as having models above the threshold, are not asked for comment. Their likely exposure is central to what the proposal means in practice.
  • Independent experts on export controls or AI policy: The two analysts comment on strategy and economics, but no one assesses whether the security rationale is credible or how the rules would be enforced. This is needed to test the official justification.
  • Timeline and process: The article calls the document a 'draft' and 'proposal', but it does not say when it would take effect, whether there is a public comment period, or who must approve it. Without that, readers cannot judge how close the measure is to becoming binding.

Highlighted Passages

“citing national security as one of the reasons”
This body wording is more qualified than the headline's 'citing national security risks', which is why the headline is a little more emphatic than the body.
“The move "can be seen as a response to Western countries posting sovereignty requirements on AI, robotics, telecommunications, and other key technologies," said Lian Jye Su”
This is the most interpretive line in the piece. It is attributed and hedged, but no Western response is given to test it, so the framing stands largely unchallenged.
“Chinese AI models including Moonshot's Kimi K3 and Alibaba's Qwen 3.8-Max are among those with parameter counts exceeding 2 trillion”
This is a factual claim about named companies' models stated without a source or a note that it is unverified, and the companies are not asked for comment.

Key Facts

  • According to the Ministry of Commerce and the Ministry of Science and Technology, China plans to restrict or ban exports of 40 technologies, covering AI, chips and solar power, in a draft update to its export catalog.
  • According to the Chinese ministries, 23 of the listed technologies are AI technologies, ranging from advanced large models to agents.
  • The draft would prohibit exports of core technologies of AI large models trained on sensitive national data, and of AI agents designed for sensitive settings such as national defense and critical information infrastructure.
  • The draft lists 12 solar-related technologies, including those used in making polysilicon, and five biomedicine technologies, including cell and gene therapy.
  • According to the commerce ministry, the move is 'necessary' and is intended to reinforce the security bottom line, prevent misuse, and keep AI under human control.
  • According to AP's reporting, the US Commerce Department asked Anthropic in June to bar foreign nationals from using some of its most powerful AI models, and those restrictions were later lifted.
  • According to Omdia analyst Lian Jye Su, the move can be seen as a response to Western sovereignty requirements on AI, robotics and telecommunications.
Read the original articlePBS NewsHour · pbs.org