How will AI Impact the English Language’s Global Dominance?

The internet is dominated by this language: English. Nearly half of all websites today are in the language of Shakespeare. However, with AI technologies now enabling seamless writing and translation across multiple languages, could these technologies undermine English’s position as the world’s global and digital lingua franca? 
 
Given the novelty of AI technologies, it’s difficult to say for certain. Where in the U.S. around 57% of the population is using AI, only 16.3% of working adults are using the technology globally. To fully understand the impact of AI on English “soft power,” its uptake would need to increase substantially. 
 
But there is serious potential. Outpacing the internet’s current translation capabilities, ChatGPT is offering more services and products across major languages. In fact, AI is now reportedly more effective at certain prompting tasks in Polish than in English. Some models now facilitate full conversations between speakers with no experience in the others’ language. With the introduction of more inclusive datasets, LLMs may eventually be able to provide comparable quality and coverage in less widely spoken languages. 
 
Will the AI revolution reverse or uphold global English language dominance? At first glance, the technology seems poised to challenge English’s primacy. China has become the second most important player in AI technologies globally. At the same time, the country is reportedly losing interest in English. The Network Readiness Index (NRI), which measures how well economies are prepared to adopt and effectively leverage digital technologies for growth and development, tells this story well. It ranks each country by technological prowess, or “network readiness.” In the NRI, China is the highest ranked upper-middle income country by a large margin. Indeed, from 2020 to 2024, China jumped 23 positions in the rankings, going from 40th to 17th. This occurred all while the country recorded a drop in global English proficiency rankings over the same period. According to Education First, or EF, which classifies non-anglophone countries by English proficiency, China’s English proficiency is ranked as “low.” Similarly, Vietnam, which only boasts “moderate” English proficiency, shows stronger NRI performance than comparable economies in its income group.
 
Looking at China and Vietnam’s trajectories one could draw that English proficiency is not necessarily a prerequisite for a country’s technological prowess. AI may intensify this trend: whereas language learners once spent years mastering the ability to write complex documents, AI can now produce them almost instantly.
 
This phenomenon is not unique to China. A 2025 study by the Centre for Economic and Policy Research (CEPR) found that demand for bilingual skills is decreasing. Demand for human translators has seen a rapid decline, and CEPR found that translation services have been widely displaced as Google Translate took hold across US counties.  
 
If we can’t determine yet how AI will affect English-language dominance, we can start by analyzing the link between English proficiency and technological readiness more broadly. Looking at the NRI, English proficiency often goes hand in hand with a country having high digital capabilities. In the 2025 NRI, countries with high English proficiency (of all income levels) had an average score of 54 compared to 50 for all countries. This is compared to an average score of 46 for Spanish speaking countries, and 33 for French speaking countries. At face value, the relationship appears to suggest that English proficiency is associated with higher network readiness, with China as an important exception.
 
However, looking at the data, this is mostly an observation: English proficiency does not cause a country to be more technologically savvy. In our regression analysis, even then, country English proficiency only explained 5% of a country’s technological prowess. More importantly, this was before per capita income was taken into account; when we factored in country income, English proficiency had no effect at all. 
Our analysis shows that China is not a one-off; widespread English proficiency is not necessary for a country to excel at digital readiness. 
 
While English might not be necessary for countries to excel in digital readiness, it might persist as the global lingua franca thanks to social media. In other words, English is often necessary to understand the latest trends on TikTok or Instagram. 

  • According to a Kuwaiti study, among 300 high schoolers, social media greatly increased reading, speaking, and understanding abilities. 
  • Another study points to TikTok as having greatly improved English proficiency among students in Vietnam. 
  • Over 97% of respondents in a study conducted in Pakistan viewed social media as a helpful vehicle for improving their English skills.

 
Although English acquired via social media is often informal, rife with slang, and that platforms such as TikTok offer little space to practice formal English writing, these platforms may be mitigating the threat AI poses to English’s status as the global lingua franca.
 
Looking ahead, AI may indeed make formal English proficiency less of a necessity for non-native English speakers in the future and may furthermore reduce incentives to master foreign languages. However, as engagement on social media suggests, English-language dominance in film, media, and culture could remain  uniquely resilient in the AI era.
 
Daniel Weinstein is a DC Fellow at the Portulans Institute.
The post How will AI Impact the English Language’s Global Dominance? appeared first on Portulans Institute.