The topic of optical modules has become a hot topic, with terms like "optical modules" and "Zhongji Innolight" trending on social media. Sources have indicated that the US government is drafting a ban to prohibit imports of new Chinese data center components, aiming to protect critical infrastructure supporting AI development.

In response, the Chinese embassy in the US urged Washington to listen to the rational voices of business communities and stop smearing Chinese companies or threatening sanctions. Analysts have weighed in to clarify the situation.

The telecommunications team at CICC noted that while short-term sentiment will undoubtedly take a hit, the long-term impact is likely minimal. They explained that optical modules primarily convert photoelectric signals and do not store business data, making their risk profile fundamentally different from active equipment. The logic cited by sources regarding FCC security standards is questionable, as is the necessity of such proposed restrictions.

Since 2020, similar narratives have emerged multiple times. Despite fears that Chinese manufacturers would lose market share to North American competitors, the opposite has occurred. Even with policies favoring US domestic supply chains, Chinese suppliers have increased their share, especially in high-speed optical modules, where they now account for 70-80% or more of orders from major overseas clients. Cutting off these suppliers would leave a gap North American firms could not fill for years, potentially slowing US data center construction.

By 2022, leading Chinese firms began establishing production in Thailand, and by 2025, companies like Yuanjie Technology and Zhongji Innolight are advancing US-based capacity. As policy patches lag, the supply chain has already evolved. In the short term, the market trades on emotion; in the medium term, it watches policy scope; in the long term, it remains about product quality, yield, and delivery. Policy may alter shipping addresses, but it cannot change who excels at making optical modules.

The telecom team at CITIC Securities echoed this view, stating the policy is unlikely to be implemented and will have limited impact. Currently, the vast majority of global high-speed optical modules come from Chinese companies, and overseas manufacturers alone cannot meet North American demand. Replacing existing modules would be prohibitively costly. Leading Chinese firms like Zhongji Innolight and Eoptolink have long established Southeast Asian production, with most North American shipments coming from these overseas facilities. Even in extreme scenarios, Chinese companies have ample room to adapt. Their technological strength, cost control, and production capacity significantly outpace overseas rivals, making a supply gap unbridgeable in the short term. Decoupling would only delay North American data center construction.

The conclusion is that the policy faces extremely high implementation difficulty and may affect sentiment temporarily, but this rumor could serve as a final stress test for the optical module sector, presenting a buying opportunity. Other analyses stress that this is merely a rumor and a draft, lacking specifics on timing or details. Many such drafts are modified or discarded after internal US regulatory debates. Historical precedents suggest these efforts often fizzle out. Given the deep penetration of Chinese optical modules in US data centers, flexible solutions exist, such as overseas production to avoid tariffs. Short-term pullbacks could be ideal moments to increase positions, as the industry's long-term trend and logic remain intact.