While large language models (LLMs) have dominated technological advancements in recent years, 2025 has introduced a major new development: AI agents
Analyzing Google Trends data, Finbold research found that search interest for ‘AI agents‘ rocketed 900% over the past year, with its sharpest increase occurring in early March 2025.
Between January 12 and February 23, the ‘trend’ score remained in the 63-71 range – albeit on an upward trajectory – before soaring to 100 by March 9. Such an increase demonstrates a major uptick in interest, as 100 is the highest possible score on Google Trends.
China dominates ‘AI agent’ search volume
Finbold research also found that most search activity is concentrated in East Asia, with the top five countries located in the region. The People’s Republic of China dominates the list, as interest within it is also at a maximum of 100.
There is a substantial gap between the first and second spots. Singapore, the second nation on the list, boasts a score of 21. The city-state is followed by Hong Kong, where the Google Trends reading shows 14.
South Korea and Taiwan both have a reading of 10 in terms of demand for ‘AI agents.’
Interestingly, the so-called collective West is absent from the top ten countries with the highest scores. Western and Central Asian nations take the second five spots: Israel, the United Arab Emirates (UAE), Pakistan, India, and Nepal.
Why China dominates ‘AI agents’ search trends
Examining the specific artificial intelligence agents and their ‘trend scores’ quickly reveals why Asian countries dominate the Google search landscape.
Specifically, the rise to prominence can mostly be attributed to the novel Manus AI agent, which was developed in China and is, at press time, available for testing only with an invitation.
Manus is, as of March 11, promising to be at least as major a disruption to the established Silicon Valley dominance as DeepSeek’s R1 model has been.
The new product is, however, different from its previous disruptive compatriot. Most currently available artificial intelligence falls roughly within the AI assistant category, meaning its operations depend on the constant influx of user prompts. Such platforms can assist with tasks but can only complete basic work.
AI agents, on the other hand, are supposed to be able to complete complex work with only minimal input, essentially replacing certain workers rather than helping them.
Though various forms of ‘AI agents’ – whether they use the technology or just the branding – exist, with stock and cryptocurrency ‘trading bots’ and price prediction tools being a particularly prominent example, China’s new platform is revolutionary, based on the available information.
Why Manus is different than existing AI platforms
Instead of being tailored to one specific task, Manus can reportedly perform a variety of tasks with minimal input and oversight. Furthermore, with its processes located in the cloud, it does not even require the end user to keep their device on for the duration of the task.
The product achieves its versatility by employing many subagents that can complete or assist in specific tasks as part of its workflow. Such a setup is unique among the platforms known at press time and represents a profound shift in the pecking order dominated by Silicon Valley.
While the fact that Manus is very new – it was launched on March 6 – and unavailable for the public means its full capabilities are unknown, reports covering its release indicate it has the ability to complete complex tasks with minimal instructions.
Allegedly, the platform can parse through documentation such as CVs and resumes, extract relevant information, and rank the candidates. It is also said that if asked to find an apartment for sale, it will not only provide the listings available online but also analyze additional factors such as neighborhood crime rates.
Should employees be anxious about AI agents?
Manus’ versatility and advanced features leave some room for concern. Western regulators have reportedly not accounted for an autonomous and multi-purpose AI agent being available by 2025, meaning there is hardly a framework for such products.
Similarly, the disposition of Chinese watchdogs toward technology remains relatively unknown, though they have apparently been focused on accommodating further advancements so far.
Similarly, the product is likely to cause much anxiety among workers around the globe, as hinted by the backlash to the December 2024 Artisan AI ‘Stop Hiring Humans’ billboard campaign.
For what it is worth, conventional wisdom has, for decades, been estimating that middle managers would be the first to go in the second machine age.
Featured image via Shutterstock
