China’s low-cost AI models, including DeepSeek, are challenging US technology companies while raising questions about data privacy, cybersecurity and national technology dependence. (Images: X)
New Delhi: China’s rapid expansion in Artificial Intelligence (AI) is opening a new front in the global technology rivalry, with low-cost AI models attracting businesses and developers while raising concerns about data privacy, national security and long-term dependence on Chinese technology. Companies such as DeepSeek have demonstrated that powerful AI systems can be developed and distributed at comparatively low costs, challenging the dominance of expensive US-based alternatives. However, critics warn that the commercial appeal of these models could come with risks involving sensitive information, data governance and the growing influence of Chinese technology providers.
A report published by the UK newspaper The Guardian argues that Western governments should look beyond the immediate financial advantages of adopting inexpensive Chinese AI. It raises questions about how user data is handled, what legal obligations companies face in China and whether widespread adoption could create strategic dependencies.
Chinese AI developers have been releasing new models at a rapid pace, offering businesses and governments alternatives to products developed by major US technology companies. For organisations looking to reduce AI deployment costs, these systems can be attractive because some models offer downloadable weights, greater customisation and the flexibility to run applications within their own infrastructure. This can give developers more control over how AI tools are integrated into business operations.
By contrast, some commercial AI services rely on paid subscriptions, usage-based pricing or proprietary systems that provide less flexibility for modification. The growing availability of lower-cost models is therefore changing the economics of AI adoption. Yet critics argue that cost alone should not determine which technology governments, financial institutions and critical infrastructure operators choose. The central question is whether organisations can adequately assess the security, privacy and geopolitical implications of relying on AI systems developed in another country.
One of the most prominent concerns involves the collection and handling of information submitted to DeepSeek's online chatbot. As highlighted in The Guardian's report, DeepSeek's published privacy terms describe the collection of user prompts, uploaded content and other information. Its online services have also disclosed that certain personal data may be processed and stored in China, subject to the applicable policy and service arrangements. Oxford University Professor of AI Michael Wooldridge, cited in the report, warned that it would not be unreasonable to consider the possibility that information entered into the chatbot could be accessible to the Chinese state. Such concerns are particularly relevant when AI tools are used to process government documents, confidential corporate material, financial information or personal data.
However, the possibility of data access should not be confused with proof that every user prompt is routinely shared with Chinese authorities. The actual risks depend on the service being used, its data practices, the applicable law and technical safeguards in place. For organisations handling sensitive information, reviewing privacy policies, data residency arrangements, contractual protections and security controls is essential before deploying any external AI service.
China's expanding AI sector is also intensifying competition with American technology companies. The appeal of lower-cost Chinese models lies in their potential to deliver useful capabilities without the same level of spending traditionally associated with advanced AI development and deployment. Open-weight systems can also allow companies to modify models for specialised applications instead of relying entirely on a vendor's hosted service.
This competition could put pressure on US AI providers to reduce prices, improve efficiency and offer more flexible products. However, the debate is not simply about Chinese AI versus American AI. It also concerns the trade-off between affordability, transparency, privacy, reliability and control over critical digital infrastructure. Organisations must assess the specific model and deployment arrangement rather than assume that a product is safe or unsafe solely because of its country of origin.
The Guardian report argues that China's low-cost AI exports could have consequences beyond conventional price competition. Chinese manufacturers have already become significant competitors in industries such as electric vehicles, batteries and other advanced technologies. The concern raised in the report is that AI could extend this competitive pressure into software, business services and digital infrastructure.
If governments and companies build essential operations around a particular foreign technology ecosystem, replacing it later could become expensive and technically difficult. Such dependence could affect access to services, data management, software updates and the ability to change suppliers. In strategically important sectors, policymakers may also need to consider whether a disruption in technology access could affect public services or national resilience. However, the extent of any future dependence will vary according to procurement decisions, local capabilities, interoperability and the availability of competing suppliers.
The debate over cheap AI is unfolding against a broader backdrop of strategic competition between Beijing and Washington. AI is increasingly important to economic productivity, scientific research, cybersecurity, military planning and industrial development. Countries that can build and deploy AI efficiently may gain advantages in both commercial markets and strategic sectors.
The Guardian report places China's AI expansion within this wider geopolitical contest, also referring to Beijing's military modernisation and tensions surrounding Taiwan, a crucial centre of advanced semiconductor manufacturing. Taiwan's importance to the global chip supply chain adds another dimension to the AI debate. Advanced processors and semiconductor manufacturing capacity underpin the development and operation of many sophisticated AI systems. Any disruption to this supply chain could have consequences for technology companies and economies worldwide.
The report also contrasts China's technology strategy with Russia's alleged use of AI to support cyber operations targeting European institutions and businesses. AI can help threat actors automate parts of their operations, analyse stolen information, generate convincing phishing messages and improve the efficiency of certain cyberattacks. At the same time, the technology can strengthen defensive capabilities by helping security teams detect suspicious activity and respond to threats.
The broader concern is that the growing spread of AI may influence international competition in ways that extend beyond traditional trade and manufacturing. The possibility that countries could become increasingly reliant on technology supplied by geopolitical rivals has prompted calls for stronger domestic capabilities, diversified supply chains and more rigorous security assessments. These risks require evidence-based evaluation, however, rather than assuming that every Chinese technology product is designed to create political dependence or facilitate state surveillance.
The concerns surrounding Chinese AI have strengthened arguments for a more cautious approach to technology procurement. Governments and businesses can reduce exposure by conducting independent security assessments, limiting the information shared with external AI services and establishing clear rules for handling confidential material.
Where appropriate, organisations can consider deploying models within controlled environments, reviewing data retention settings and requiring suppliers to provide clear information about their security practices. Policymakers can also encourage domestic AI research, support European and other international suppliers, improve access to computing infrastructure and develop consistent standards for privacy and cybersecurity. Such measures would allow organisations to benefit from AI competition while reducing the risks associated with concentrating essential systems in the hands of a small number of providers.
China's low-cost AI models are challenging established assumptions about the price of advanced technology. Their availability could expand access to AI, lower operating costs and intensify competition among global developers. At the same time, questions about data protection, regulatory oversight and technological dependence deserve serious scrutiny, particularly when AI is used in government operations or sensitive industries. The policy challenge is not simply whether to accept or reject Chinese AI. It is how to evaluate individual systems, protect sensitive information and maintain technological resilience without sacrificing the economic benefits of innovation.
As AI becomes more deeply integrated into everyday business and public services, decisions made today about suppliers, data and infrastructure could shape countries' technological independence for years to come.
