CITIC Securities: AI Disrupts the Narrative of U.S. Internet Stocks, Short-term Overinterpretation

CITIC Securities’ research report states that the AI narrative overturning the US stock market’s internet story has been overplayed in the short term. In consumer scenarios, the incremental gains from AI experiences are limited; AI replacement faces cost constraints; and model companies themselves have inherent capability boundaries. Therefore, AI is more likely to be a cooperative relationship rather than a replacement relationship with existing internet platforms, and some high-quality companies have been clearly sold off unfairly. CITIC Securities recommends focusing on companies that have AI-era competitive barriers such as links to the physical world, strong network effects, accumulated data and algorithms, and high-quality content IP, while also finding directions in which demand expands as AI penetrates.

The full text is as follows

Outlook | The US Internet Investment Logic in the Context of AI Agents

The AI narrative overturning the US stock market’s internet story has been overplayed in the short term. In consumer scenarios, the incremental gains from AI experiences are limited; AI replacement faces cost constraints; and model companies themselves have inherent capability boundaries. Therefore, AI is more likely to be a cooperative relationship rather than a replacement relationship with existing internet platforms, and some high-quality companies have been clearly sold off unfairly. We recommend focusing on companies that have AI-era competitive barriers such as links to the physical world, strong network effects, accumulated data and algorithms, and high-quality content IP, while also finding directions in which demand expands as AI penetrates.

▍ AI is advancing rapidly, but it is not all-powerful.

Over the past two years, real breakthroughs have emerged in three major bottlenecks that have constrained the large-scale application of Agent models: model reliability, inference costs, and ecosystem interoperability. This has driven deeper, more productized deployment of AI in core internet scenarios such as e-commerce, advertising, and content. At the same time, market panic about existing internet platforms being overturned has been growing day by day. We believe the actual impact of AI is milder than what the market expects, and the limiting factors come from three levels:

1）In consumer internet scenarios, the incremental gain in AI experience has been overestimated. Unlike in B-end scenarios, in C-end decision chains for shopping, commuting, dining, and the like, there is already limited repetitive work that AI can significantly improve. Meanwhile, core issues such as information completeness, payment security, and platform reliability have not yet been effectively solved under existing AI frameworks;

2）Constraints from the overall cost of AI replacement mean it is difficult for existing platforms to be comprehensively rebuilt. After subscription fees, maintenance investments, interface integrations, and other implicit costs are added, the cost-effectiveness becomes significantly worse than directly integrating into mature ecosystems;

3）The capability boundaries of large-model companies themselves and the diseconomies of scale phenomenon determine that the probability that large-model companies cooperate with vertical platforms rather than replace them is higher. The profitability pressure after going public will also force them to focus on their core capabilities. Therefore, we suggest objectively and calmly assessing AI’s impact on the internet sector. What AI brings is structural differentiation rather than a systemic overthrow.

▍ Finding high-certainty, structurally incremental opportunities in the AI era.

At the level of impact, on the one hand, AI brings quantifiable business incremental gains to internet platforms through three paths: enhancing user experience, optimizing operational efficiency, and opening up new business scenario opportunities. On the other hand, the replacement effect of AI-native products on traffic entry points such as traditional search and content distribution is starting to become apparent.

At the level of moats, against the backdrop of the mixed picture mentioned above, whether a platform can sustain its competitive moat in the AI era is the key to judging investment value. We believe that companies that have links to the physical world, strong network effects, accumulated data and algorithms, and high-quality content IP will stand out in the AI wave.

At the level of opportunities, through traffic diversion and disintermediation, a substantial portion of commercial value will be captured at the model layer rather than flowing to existing platforms. However, some internet platforms still have opportunities to share in the AI dividend. Their common feature is that they are outside the scope of direct AI replacement, but service demand will expand systematically as AI penetration rates increase. Advertising platforms, streaming media, cloud computing, and others are typical examples. Based on the two factors of competitive moats and AI opportunities, we conduct integrated quantitative and qualitative analysis, and we divide the key US internet companies we cover into four quadrants: beneficiaries, divergence points, safe harbors, and losers, among which some companies are more likely to benefit from this round of the AI wave.

▍ Risk factors:

Risk that AI progress far exceeds expectations and therefore increases the severity of the shock; risk that AI infrastructure investment is too heavy and the uncertainty of returns is high; risk that AI disrupts the content ecosystem; risk of intensified market competition; risks such as tighter regulatory policies related to data and platform operations.

▍ Investment strategy:

We believe that AI has achieved breakthroughs in key bottlenecks and is accelerating deployment. However, the narrative that AI will overturn the internet is overly pessimistic, mainly because in consumer scenarios the incremental gains from AI experience are limited, AI replacement faces cost constraints, and large-model companies have capability boundaries. We recommend focusing on companies that have AI-era competitive barriers such as links to the physical world, strong network effects, accumulated data and algorithms, and high-quality content IP, while also finding directions in which demand expands as AI penetrates.

（Source: Jiemian News）