Game-Theoretic Framework for International AI Cooperation:
Comparative Case Study of China and India
Yichen Wu
a
SSFBC- The High School Affiliated to Shaanxi Normal University, Xi’an, Shaanxi, 710061, China
Keywords: Global AI Development Inequality, Game Theory, Global AI Cooperation, China, India.
Abstract: In this era of rapid AI development, many countries and industries have benefited from this new technology.
Due to the emergence of AI, many researchers have researched policies of AI development and forecasted the
future development tendency of AI. However, few have addressed global AI inequality through mathematical
and economic lenses, and policy conflicts of AI between different countries as well. To bridge this gap, this
paper employs a game-theory model to analyze the lack of global cooperation in AI development. At present,
it provides suggestions to solve this inequality situation. China and India are employed as representatives of
dominant AI countries and AI developing countries, respectively. AI technology in China has developed
quickly in recent years because China has overcome technical obstacles, exemplified by breakthroughs like
Deepseek, China has successfully ascended to the rank of dominant AI countries. India is an AI-developing
country with many skilled people, but India fails to retain them domestically, and India does not have enough
technology and money to do further research in AI development. The article summarizes both external and
internal reasons for the global AI development inequality problem. Then, suggestions from these two aspects
based on the mathematical calculation in the game theory model and reasons for the problem that have been
analyzed will be provided. Finally, the author indicates the significance and limitations of this research.
1 INTRODUCTION
The rapid advancement of Artificial Intelligence (AI)
has transformed industries such as healthcare, law,
finance, and education, demonstrating significant
scalability (Hine & Floridi, 2024). However, AI
adoption remains uneven globally. Post-pandemic
data reveals a 28% growth in AI investment in
developed nations compared to just 9% in developing
economies, exacerbating global AI inequality.
The U.S. dominates AI innovation, exemplified
by ChatGPT. Studies show GPT-3 achieves an IQ of
~150 (99.9th percentile), while GPT-3.5 excels in
professional certification exams (Ray, 2023). Its
applications span finance, healthcare, and media,
enhancing efficiency in developed economies (Nazir
& Wang, 2023).
China has also achieved breakthroughs, launching
Deepseek—a leading Large Language Model (LLM).
Comparative studies rank Deepseek above Claude,
Gemini, GPT, and Llama in text classification
accuracy and cost efficiency (Gao, et al., 2025). This
a
https://orcid.org/ 0009-0006-0859-2401
progress stems from China’s strategic policies,
including the Next Generation AI Development Plan
($15B allocated for international collaborations) and
a focus on technical education. Despite these efforts,
China still faces challenges in overcoming
technological monopolies from AI-leading nations.
Many developing nations—such as India,
Vietnam, and the Philippines—aspire to advance AI
but face severe challenges, including limited GDP,
skill shortages, and digital infrastructure gaps. This
raises a critical question: Is international cooperation
essential for equitable AI development? While
leading AI nations (e.g., U.S., China) could share
expertise, they often prioritize monopolizing core
technologies to maintain competitive advantages.
AI’s transformative potential across industries
incentivizes such protectionism. For instance,
Foffano et al. highlight regional alliances like the
Nordic-Baltic AI collaboration, contrasting with the
U.S.-China "tech decoupling" (Foffano, et al., 2023).
Schmidt notes that military AI applications further
258
Wu, Y.
Game-Theoretic Framework for International AI Cooperation: Comparative Case Study of China and India.
DOI: 10.5220/0013823200004708
Paper published under CC license (CC BY-NC-ND 4.0)
In Proceedings of the 2nd International Conference on Innovations in Applied Mathematics, Physics, and Astronomy (IAMPA 2025), pages 258-266
ISBN: 978-989-758-774-0
Proceedings Copyright © 2025 by SCITEPRESS – Science and Technology Publications, Lda.

complicate bilateral trust, forcing selective
cooperation in shared interests (Schmidt, 2022).
For economically constrained countries,
independent AI development is costly and slow.
Vietnam’s higher education sector, for example,
struggles with inadequate technical equipment, yet
digital transformation requires prohibitive
investments (Quy et al., 2023). Such nations thus seek
partnerships to access funding and technology, but
power asymmetries persist.
Key Evidence of Inequality: Monopoly vs.
Cooperation: Dominant nations restrict critical AI
transfers (e.g., U.S. export controls on chips for AI
training). Regional Disparities: Nordic-Baltic
collaboration vs. developing nations’ isolation
(Foffano, et al., 2023). Economic Barriers: Vietnam’s
digital education hurdles reflect systemic
underinvestment (Quy et al., 2023).
These cases underscore how AI inequality stems
from structural imbalances (resources, policy) and
geopolitical competition (U.S.-China rivalry).
Addressing this requires rethinking cooperation
frameworks beyond zero-sum dynamics.
Most researchers overview the AI development
situation in different countries by researching policies
promulgated by AI development in those countries
and forecasting the future development tendency of
AI. Additionally, a number of scholars focus on
investigating the policy conflicts of AI among several
countries. However, few researchers have analyzed
the competitive or cooperative relationships between
powerful countries and relatively poorly developed
countries in AI technology by utilizing mathematical
or economic models.
Building on these documented disparities, game
theory provides an analytical framework to model the
strategic interactions between China and India. This
paper applies game theory as the basic theory, China
as the powerful AI country, and India as the relative
lack of developed country in AI fields as examples to
analyze different payoffs that can bring to the
countries based on different decisions. Therefore, the
article formulates the suitable development patterns
in AI technology in these two countries, which can
maximize payoffs and commit to settling the
inequality problem in AI fields.
2 DIVERGENT PATHWAYS: AI
DEVELOPMENT IN CHINA
AND INDIA
Based on the problems of global AI development
inequality, this article aims to model strategic
interactions through game-theoretic frameworks to
derive optimal cooperation mechanisms by
employing the game theory model and anticipating
the payoffs of each country to observe the
competitive or cooperative relationship between
powerful AI countries and AI developing countries.
This essay chooses China and India as examples to
represent these two types of countries and aims to
discover their choices in terms of encountering AI
development inequality problems.
When generally reviewing the AI development of
China in the past decade, it was a process from
accumulation to rapid progress. The tide of AI
development in China appeared in 2010; with support
from government policies, AI development became a
national strategy in China five years later (Qu & Li,
2022). Until the second half of 2022, the core AI
industry scale in China had closed up to the leading
ranks of AI technology around the world, which was
more than 400 billion CNY and had over 3000
companies (Qu & Li, 2022). Tencent Holdings
Limited, Baidu, Inc, and Huawei Technologies Co.,
Ltd are examples of companies that have
continuously made progress in AI development. They
helped China overcome technical difficulties and
applied AI to medical treatment, education, and other
industries. In the aspect of international AI
development policies, China has always had a
positive attitude as an important international force of
AI technology in recent years. China appeals to
prevent digital hegemonism and promotes AI
globalization by coming up with the “Global AI
Governance Initiative” to help other countries
develop AI (Li & Fan, 2025). Having established
China’s AI leadership position, this study now
examines India’s contrasting development trajectory.
As a developing nation, India recognizes AI's
strategic importance across industries and defense,
mirroring global leaders' urgency. With aspirations to
become a tech power, India views AI advancement as
critical for competitiveness. However, it faces
challenges in talent retention and core technology
access despite its software outsourcing strengths.
India has certain advantages in AI development.
Initially, India is one of the dominant countries that
provides software outsourcing services with a large
population and relatively low human cost (Wei, 2024).
Additionally, the large population provides India with
a number of potential educable technical human
resources. Apart from these strengths, India
encounters a dilemma as well. The lack of skilled
people and the need to break through technical
difficulties in chip production are two main
components of the dilemma. India’s government has
Game-Theoretic Framework for International AI Cooperation: Comparative Case Study of China and India
259

issued some supportive policies and funding to
resolve these problems since 2018, and it has a strong
incentive to cooperate with those powerful AI
countries. Thus, India had some cooperative projects
in AI development with the United States and Japan.
In the past few years, India and China have
collaborated on AI. They co-founded “Sino-Indian
Digital Collaboration Opportunity Plaza” and China
supported several initially established Indian
companies for operation (Wei, 2024). However, India
and China started to unhook in recent years, which
means the AI cooperation situation between the two
countries become more ambiguous (Wei, 2024).
Although India has cooperated with some
powerful AI countries, these countries will not
provide India with the core technology. Therefore,
this article will concentrate on whether India should
strive to learn the core technology from these
dominant countries in AI technology to collaborate
with them or subsidize more on independent research
and cultivating skilled people.
3 STRATEGIC INTERACTION
ANALYSIS
3.1 Structural Inequality in AI
Ecosystems
3.1.1 Global Disparities
With the quick development of AI technology in
some countries, global AI development presents
several problems that increase inequality. One of the
problems is the unequal data sources. More than 90%
of AI model training data are from countries with the
English language, which is a factor that reduces the
language and cultural diversity in developing
countries (Kallus, 2023). Besides, some powerful AI
countries become monopolies in the AI field to
consolidate their dominant places in AI technology
and compete with other dominant countries.
Therefore, the difficulties for AI-developing
countries to make breakthroughs are high, and the
panic of those AI-developing countries will increase
due to a lack of technology and skilled people.
Additionally, the imbalance in computing power
distribution creates an insurmountable barrier as well.
Geographical factors reinforce the hierarchies in
computing power. The top 1% of institutions control
85% of the resources of global AI development
computing in the AI field, which seriously increases
the unequal gap (Ahmed & Wahed, 2023). Last but
not least, the mobility and emigration of skilled
people is an uncontrollable problem for AI-
developing countries. These AI-developing countries
do not have enough funding to invest in the training
of talented people and high-level education. They
cannot prevent these people from moving to those AI-
developed countries to find better job opportunities,
with India losing 42% of top AI graduates annually to
4.7x higher salaries abroad. Therefore, the potential
for AI development in these AI-developing countries
will be limited.
3.1.2 India’s Developmental Constraints
As an AI-developing country, India encounters
several problems that impede the development of AI.
The inconsistent digital infrastructure is one of these
problems. Because of the incomplete coverage (Only
34% of rural areas have >10Mbps internet vs China’s
89%) and unreliable connectivity of the internet in
India, especially in some rural areas, the scalability of
AI solutions will be limited (Sircar & Singh, 2023).
These limitations caused by infrastructure and limited
resources bring a huge problem to India’s AI
development. In addition, the quality of massive data
utilized to train AI does not meet the standard, and its
usability is suspicious. This postponement and lack of
high-quality information are setbacks that lower the
speed of AI development in India. Moreover, the
brain drain problem is a common phenomenon in AI-
developing countries, as well as in India. India spends
a large amount of money on producing STEM
graduates, but the lack of advanced AI institutions
and technical jobs with high salaries leads to more
skilled people moving to Western countries to work
and live (Patel & Khanna, 2023). This is the main
reason that fewer domestically skilled people work in
India and that the Indian AI development speed has
not increased, although it has invested largely in
captivating skilled people. The last problem is that the
Indian government has not made clear policies in AI
development; the lack of clear guidelines in privacy
safety, ethics, and industries leads to ambiguous AI
development policies and low improvement speed
(Verma & Rao, 2022). This disorder caused by the
government represents the lack of consciousness of
AI development in India and misgoverning in the AI
field, which is the main reason for its poor
development.
3.2 Game-Theoretic Modeling
3.2.1 Stackelberg Game Framework
In the area of AI development, because of the
differences in technology development levels in
IAMPA 2025 - The International Conference on Innovations in Applied Mathematics, Physics, and Astronomy
260

different countries, the movements of talented people,
and other exogenous variables, the global
development of AI technology is becoming more and
more unequal. Although many countries have issued
some external policies related to AI development
based on their national conditions, the general AI
development trend is not friendly to those countries
with technical obstacles. Some powerful AI countries
concentrate more on competing with other dominant
countries in the AI field, which means they naturally
form technical monopolies and indirectly cause less
concentrating problems in those AI developing
countries from the international view. In order to
discover the reasons for the formation of the present
situation of global AI development and provide
potential policies to solve the unequal problem, this
article here employs the mixed strategy in game
theory to analyze and utilize China as the AI-
developed country, India as the AI developing
country as examples.
Game theory is a theory based on mathematical
methodology that studies how multiple individuals or
teams make decisions and how these decisions reach
equilibrium when the individuals or teams mutually
influence each other. It can provide an equilibrium
solution framework for and help make decisions
based on the rationality of opponents (Abedian et al.,
2022). Two typical strategies are in the game theory.
Pure strategy is a type of strategy in the complete
information game in which each player can only
choose one specific strategy under the given
information. This article applies another strategy
known as the mixed strategy. Mixed strategy is
defined as a probability distribution over the set of
pure strategies available to a player. The player
randomly chooses among different pure strategies
based on different probabilities. The probabilities are
calculated based on the expected payoff of different
choices. This article applies mixed strategies because
of the complicated real-world situation of global AI
development. A mixed strategy can analyze the
strategies of the two countries with more
comprehensive consideration when both countries do
not have their dominant strategy, and there is no pure
strategy Nash equilibrium in the game. Here, the
author employs Table 1 to represent the specific
payoff when China and India face different global AI
development policies.
Table 1: The payoff matrix of China and India
China (AI developed
country)
India (AI
developing country)
Cooperation
in AI
development
Monopoly in
the AI field
Cooperation in AI
developmen
t
e, d h, b
Individually research and
develop AI
f, a g, c
3.2.2 General Payoff Introduction
In Table 1 above, China and India represent an AI-
developed country and an AI-developing country,
respectively. Table 1 assumes two options for each
country. For China, the options are cooperating with
India in AI development or forming a monopoly in
the AI field. For India, the options are cooperating
with China to increase the speed of AI development
or doing the research and development individually to
break the technical obstacles. Therefore, four pairs of
payoffs are presented in Table 1 based on different
option patterns between the two countries. In order to
demonstrate each payoff clearly, the author utilizes
the small letters s ‘a’ to ‘h’ to represent the eight
different payoffs for the two countries when facing
different options, and they can have specific values.
If both countries choose to cooperate in AI
development, the pair payoff for them is（e, d）, ‘e’
is for India, and ‘d’ is for China. Based on the same
principle, if India is willing to cooperate while China
chooses to be a monopoly in the AI field, the pair of
payoffs is (h, b). Additionally, if India chooses to
develop AI technology independently and China is
willing to cooperate, the pair of payoffs is（f, a).
Finally, if India continues to choose research
independently and China chooses to become a
monopoly in AI technology, the pair of payoffs is（g,
c).
In this model, the value of a＞b＞c＞d＞e＞f=g
＞h. In general, neither country has any dominant
strategies in this model, which means a mixed
strategy can be utilized to indicate the expected
payoff for each country based on different
probabilities of two opportunities that another
country will choose. If India chooses to cooperate
with China in AI development, China can receive a
higher payoff through being a monopoly as well. And
if India chooses to research individually, the better
response for China is choosing cooperation because
‘b’ ＞ ‘d’ and ‘a’ ＞ ‘c’. When China chooses to
cooperate, India can receive a higher payoff through
choosing cooperation. And if China chooses to
become a monopoly in the AI field, independent
Game-Theoretic Framework for International AI Cooperation: Comparative Case Study of China and India
261

research to develop AI technology can bring a higher
payoff to India because ‘e’ ＞ ‘f’, ‘g’ ＞ ‘h’.
In general, all the payoffs of China are higher than
all the payoffs of India because even if China chooses
to cooperate with India in AI development and share
some technology with India, China still has a larger
market share, a more developed technology
foundation, and the number of skilled people than
India. The payoff for China will be lower when the
two countries cooperate, but the leading place of
China in the AI field cannot be simply shaken.
Additionally, suppose China chooses to be a
monopoly in the AI field. In that case, the payoff of
China will be apparently higher than any payoff that
India can receive among all options due to the
technical obstacles. Therefore, in this model, China
always has advantages in getting higher payoffs than
India, which is the reason that a＞b＞c＞d＞e＞f＝
g＞h in the payoff matrix.
3.2.3 Specific Payoff Analyzing for China
For China, if India chooses to develop AI technology
independently and China chooses to cooperate, China
can receive the highest payoff among the four payoffs
of China, which is the value “a”. This is because
choosing cooperation can reflect the open and
inclusive policies in the diplomacy of China and it is
a great chance to erect the helpful international image
of China. This positive international image can bring
many further benefits to China in terms of
international relationships and economic trade with
other countries. The action of being willing to
cooperate only causes a small amount of cost for
China, which can be ignored. Additionally, because
India is willing to develop AI technology
independently, China will not sell the AI technology
to India, which means China can maintain its
competitiveness in the international AI market.
However, if India’s research is successful in some
areas, the competitiveness and market share of the AI
industry will increase. This private benefit India
receives will threaten China’s market share to a small
degree. Nevertheless, the negative effects that this
option pattern has brought to China are much less
than the benefits that China can receive. Therefore,
China is willing to choose to cooperate when India
independently researches, and China’s payoff is
highest in this option pattern.
The second highest payoff for China occurs when
India chooses to cooperate, and China chooses to be
a monopoly. Under this option pattern, China can
maintain its competitiveness in the AI field in the
global AI market due to the unsuccessful cooperation
with India because China will not sell the exclusive
technology to India. In addition, it is time-consuming
for India to find cooperation and persuade China. The
money and time spent on negotiating and
transportation will distract India’s attention from
developing AI, which means the opportunity cost of
finding cooperation is high. India can receive a higher
payoff by utilizing the funding and time in directly
choosing to research individually. Since India has
wasted time and money, the achievement it can
receive will be lower, which means it can only share
a small amount of market share. Therefore, the
negative influences brought to the Chinese payoff
will be small compared with the market
competitiveness China can maintain. However, China
cannot receive further benefits from the international
image under this option pattern. This is the reason that
the value of “b” is smaller than“a” in this model.
When both China and India are not willing to
cooperate, China can receive the third highest payoff.
Under this option pattern, India can share more
market share because it spends money and time on
independent research directly, which means India
saves on opportunity costs and will have more
breakthroughs. Besides, due to the unsuccessful
cooperation, China can still maintain its
competitiveness in the global AI market and compete
with other dominant AI countries. Therefore, the
payoff that China receives will be lower than when
India has a small market share, which means the value
of “c” is smaller than “b”.
The last payoff of China occurs when both
countries cooperate successfully. Under this option
pattern, China can receive further benefits from its
positive international image. China will also receive
a patent fee when it sells AI technology to India.
However, due to the successful cooperation, the speed
of AI development in India will become much faster
than before, and India will make more breakthroughs
in the AI field, which means it can take away a large
amount of market share from China because of the
same technology they have.
This separation in the global market share brings
a huge loss of Chinese payoff, and the loss is much
higher than the benefits China can receive from
cooperation. Moreover, China cannot maintain its
competitiveness compared with other dominant AI
countries because India has the same technology as
China. Therefore, the payoff of China is the lowest
under this option pattern, which explains the reasons
that the value of “d” is the lowest among “a” “b” “c”
and “d”.
IAMPA 2025 - The International Conference on Innovations in Applied Mathematics, Physics, and Astronomy
262

3.2.4 Specific Payoff Analyzing for India
For India, the highest payoff occurs when both
countries choose to cooperate. With the help of China,
India can develop AI technology much more quickly
than before. And the signs of progress will bring
several private benefits to other domestic industries.
For example, the quality of education and medical
treatment will increase due to the utilization of AI,
which means that domestic living standards will
increase and inequality problems will be reduced.
Furthermore, the external benefits India can receive
are large as well. India can gain a large market share
because of technological breakthroughs. This market
share can widen India’s international AI market and
facilitate India’s receiving higher profits from global
trades related to the AI industry. However, the patent
fee should be considered India’s main cost under this
option pattern, but compared with the benefits India
can receive from cooperation, the cost is relatively
small. Therefore, India’s payoff is highest when the
two countries cooperate successfully.
If India directly chooses to develop AI technology
independently, the payoff for India will be the same
whether China chooses to cooperate or be a monopoly.
Under these option patterns, India does not need to
pay for the patent fee, but the progress it can make
will be less than cooperating with China. Thus, the
market share and private benefits India can receive
are smaller than in the case of cooperation. Generally,
although the benefits of India are much smaller than
before, the little progress can bring a small scale of
beneficial influence to India, which means the
payoffs are not the highest, but not too bad. This is
the reason that e＞f＝g in the payoff matrix.
India receives the lowest payoff when China
chooses to be a monopoly, and India is willing to
cooperate. India will put effort and money into
facilitating the cooperation, thereby distracting
India’s attention to AI development. After wasting
time and money, Indian initial funding for AI
development will decrease due to the opportunity cost,
which means India can make less progress than
directly concentrating on independent research.
Therefore, the market share and private benefits India
can receive will be much lower than the benefits it can
get from individual research. And with the cost of
facilitating cooperation, India’s payoff is the lowest
among the four payoffs. Thus, “h” is the lowest value
among all payoffs in the matrix.
3.2.5 Mixed Strategy Equilibrium
The mixed strategy Nash equilibrium requires the
following: the probability of China choosing
cooperation is ‘p’. To choose to become a monopoly
is ‘1-p’. The probability of India choosing
cooperation is ‘q’, and choosing to conduct individual
research is ‘1-q’. Based on the relationship of the size
of the eight payoffs, the author adds eight specific
values of the payoffs to make further descriptions of
choices between the two countries.
Table 2: The payoff matrix of China and India with
specific values and probabilities
Probabilit
y
p 1-p
China
India
Cooperation
in AI
development
Monopoly
in the AI
field
q Cooperation in
AI development
5, 6 3, 8
1-q Individually
research and
develop AI
4, 10 4, 7
When the probability of China choosing to
cooperate is “p”, to choose to be a monopoly is “1-p”,
the total expected payoff of India can be written as:
𝔼

π

=q

5p + 3

1−p


+

1−q


4p + 4

1−p


=2pq−q+4 (1)
Then, the relationship between E(π) and “p” can
be demonstrated as:
𝔼




=2𝑞 (2)
The “q” is a number between 0 and 1, so 2q＞0,
which means “p” and 𝔼

𝜋

have a positive
relationship. When the value of “p” increases, the
value of 𝔼

𝜋

increases. And if “q” is greater, when
“p” increases, 𝔼

𝜋

increases with a faster speed.
Therefore, India is willing to see China have a higher
probability of tending to cooperate, which can lead to
a rise in the total expected payoff of India.
When the probability of India choosing to
cooperate is “q”, to choose to research and develop
AI technology independently is “1-q”, the total
expected payoff of China can be written as:
𝔼

π

=p

6q + 10

1−q


+

1−p


8q + 7

1−q


=3p−5pq+q+7 (3)
Then, the values of “p” and “q” can be calculated
by utilizing the following equation:
𝔼

 


= 1 − 5𝑝 = 0 𝑆𝑜, 𝑝 =


(4)
𝔼

 


= 1 − 5𝑝 = 0 𝑆𝑜, 𝑝 =


(5)
𝔼

 


= 1 − 5𝑝 = 0 𝑆𝑜, 𝑝 =


(6)
Game-Theoretic Framework for International AI Cooperation: Comparative Case Study of China and India
263

Therefore, when p=


,q=


, China’s expected
payoff will be maximized, and China is not willing to
see any changes in the probability that India will
choose between different options.
Thus, from the calculation above, India wants to
promote the value of “p” to increase, which can lead
to an increase in the value of “q” to increase Indian
total expected payoff. However, if “p” is greater than


, the relationship between the value of “p” and the
total expected payoff of China will be negative, which
means
𝔼

 


<0. Then, China hopes the value
of “q” to decrease to increase the total expected
payoff of China. Therefore, there is a conflict
between the choices of the two countries. These
mathematical procedures explain in detail the reasons
why it is difficult for dominant AI countries to
cooperate with AI-developing countries. If they are
willing to find a balanced solution when encountering
this game, both of them should make a concession.
4 SUGGESTIONS
4.1 Framing the Solution: A Balanced
Approach to AI Inequality
As this article mentioned before, the global AI
development inequality problem is caused by both
external global reasons and internal reasons from the
AI developing countries. Therefore, to resolve these
problems, both the global and inner factors of
developing AI countries should be considered. The
suggestions will try to find the balance point of AI
development between dominant AI countries and AI
developing countries. In the complex global
landscape, trade-offs are inevitable, and no country
can unilaterally maximize benefits without
considering broader implications. Any choices a
country makes should consider some practical factors,
such as feasibility, morality, international image, and
the concept of community with a shared future for
mankind. Therefore, in the field of AI development,
no country can be alone. By considering the
complicated reasons, many dominant AI countries are
willing to help AI-developing countries to a certain
degree. However, this willingness has not been well
implemented due to the loss of payoffs in AI-
dominant countries. Thus, the suggestions will
provide more cooperation potential for the two types
of countries.
4.2 Global Strategies: Reducing AI
Disparities
Unequal data distribution and language barriers
hinder AI-developing countries' access to quality
training data, which means their AI products cannot
be trained with updated information. To solve this
problem, firstly, the data should be translated into
different languages to increase language and culture
diversity in the AI database for those AI developing
countries and help them to train their AI products
conveniently. Additionally, it promotes the utilization
of open-access AI models in AI-developing countries.
Then, they can have more computing resources
without building expensive data infrastructure and
spending money on collecting updated data.
Moreover, the dominant AI countries should
concentrate less on competition with other dominant
countries and focus more on solving global AI
technology inequality. If they all help AI-developing
countries, their competitiveness will remain the same.
At the same time, they can have a better global
reputation and image. To foster equitable AI growth,
both North-South Cooperation and South-South
Cooperation can be utilized. For instance, the U.S.
and India’s AI collaboration demonstrates how
North-South cooperation can be mutually beneficial.
China is a developing country with advanced AI
technology, and it has several AI cooperations as well.
As the game theory model mentioned before, China
and India have a trade-off in whether they should
cooperate. By considering the payoffs of both
countries, when China cooperates with India in some
fundamental AI technology, both countries can
benefit. This is because China can receive a better
global image through cooperation. It does not sell the
most advanced and exclusive technology to India, so
China can maintain its competitiveness. At the same
time, India can receive some assistance from China to
increase its AI development speed, which will benefit
India. Therefore, other dominant AI countries and AI
developing countries can utilize similar methods
between China and India to solve the unequal
problem. Last but not least, applying for international
investment to construct some high-power computing
instruments in AI developing countries to increase
their computing power and prevent the monopoly in
the top computing institutions to reduce inequality.
4.3 National Strategies: Building AI
Capacity from Within
A key domestic challenge is the lack of cohesive AI
policies, leading to fragmented efforts." Therefore,
IAMPA 2025 - The International Conference on Innovations in Applied Mathematics, Physics, and Astronomy
264

the most important thing for the governments of these
AI-developing countries is to recognize the
importance of AI development and establish clear
policies to facilitate AI technology improvement. In
addition, although some AI-developing countries
have trained some skilled people, the emigration rate
is high due to low incomes in their domestic countries.
Thus, properly increasing the income for those
technically skilled people and increasing subsidies to
those AI development projects can attract more
domestic skilled people to help local AI development.
Moreover, some infrastructures should be improved
as well. Governments should leverage international
partnerships to secure funding for critical
infrastructure, such as internet connectivity and
power supply in underserved regions.
5 CONCLUSION
This study examines global AI inequality through a
game theory analysis of China-India dynamics. It
identifies both external factors (like data/resource
disparities) and internal challenges (particularly
India's policy fragmentation and brain drain). The
model frames China as representing developed AI
nations and India as developing ones, using payoff
matrices to quantify cooperation barriers. Results
reveal how short-term competition often outweighs
collective benefits, perpetuating inequality.
Solutions require multilateral action: Developed
nations should provide foundational AI tools (like
open-access models) while maintaining core IP,
creating balanced cooperation. International
investments must expand computing infrastructure to
break institutional monopolies. For developing
countries, three priorities emerge: (1) implementing
coherent AI policies, (2) incentivizing skilled
workforce retention through competitive
compensation, and (3) upgrading digital
infrastructure - especially in rural areas - to enable
equitable AI adoption.
This article provides suggestions for AI
development for developing countries by applying
the case study of India. This article utilizes the game
theory model to explain the detailed decision-making
process when dominant AI countries and AI
developing countries encounter trade-offs and
different option patterns. These mathematically
supportive procedures indicate the balance between
the two types of countries, and the suggestions
provide more available methods to improve the
development potential of AI technology for those
developing AI countries. This explanation fills the
gap in the research in utilizing mathematical models
and economic methods to solve the inequality
problem of global AI development, which will benefit
local residents and industries in AI-developing
countries.
Since the article applies China and India to
represent dominant AI countries and AI developing
countries as a case study, the suggestions may not be
suitable for all countries facing this problem. The
actual national situations are different in different
countries, so specific solutions for the inequality
problem are different. Therefore, only utilizing two
countries as examples may have a halo effect on the
solutions. Additionally, the values added to eight
payoffs in the matrix are only based on the theory.
They can be more valid by considering factors like
national GDP and the specific market shares they
received. Future research should establish a typology
of AI-developing nations to calibrate policy
recommendations.
REFERENCES
Abedian, M., Amindoust, A., Maddahi, R., & Jouzdani, J.,
2022. A game theory approach to selecting marketing-
mix strategies. Journal of Advances in Management
Research, 19(1), 139-158.
Ahmed, N., & Wahed, M., 2023. The De-democratization
of AI: Deep Learning and the Compute Divide. Journal
of Artificial Intelligence Research, 76, 135-158.
Foffano, F., Scantamburlo, T., & Cortés, A., 2023.
Investing in AI for social good: an analysis of European
national strategies. AI & society, 38(2), 479-500.
Gao, T., Jin, J., Ke, Z. T., & Moryoussef, G., 2025. A
Comparison of DeepSeek and Other LLMs. arXiv
preprint arXiv:2502.03688.
Hine, E., & Floridi, L., 2024. Artificial intelligence with
American values and Chinese characteristics: a
comparative analysis of American and Chinese
governmental AI policies. Ai & Society, 39(1), 257-
278.
Kallus, N., 2023. Treatment effect risk: Bounds and
inference. Management Science, 69(8), 4579-4590.
Li, J. G. & Fan, C., 2025. Core Issues of Global Generative
Artificial Intelligence policy Norms and their impact on
the world. International Review, (01), 130-156.
Nazir, A., & Wang, Z., 2023. A comprehensive survey of
ChatGPT: advancements, applications, prospects, and
challenges. Meta-radiology, 1(2), 100022.
Patel, M., & Khanna, T., 2023. The AI Talent Exodus: Why
India Struggles to Retain Its Best Minds. International
Journal of Educational Technology in Higher
Education, 20(1), 1-18.
Qu, Z. F. & Li, Z. H., 2022. Leapfrog development of AI in
China: From start-up tide to empowering thousands of
industries. China Business Journal, C01.
Game-Theoretic Framework for International AI Cooperation: Comparative Case Study of China and India
265

Quy, V. K., Thanh, B. T., Chehri, A., Linh, D. M., & Tuan,
D. A., 2023. AI and digital transformation in higher
education: Vision and approach of a specific university
in Vietnam. Sustainability, 15(14), 11093.
Ray, P. P., 2023. ChatGPT: A comprehensive review on
background, applications, key challenges, bias, ethics,
limitations and future scope. Internet of Things and
Cyber-Physical Systems, 3, 121-154.
Schmidt, E., 2022. AI, great power competition & national
security. Daedalus, 151(2), 288-298.
Sircar, I., & Singh, P., 2023. The Digital Divide and AI in
India: Challenges of Inclusive Development. Journal of
Asian Technology Policy, 8(1), 45-62.
Verma, S., & Rao, K., 2022. Regulating AI in India: Policy
Gaps and Institutional Challenges. AI & Society, 37(3),
987-1002.
Wei, Y. F., 2024. Building a Digital Superpower: An
Analysis of India’s National Artificial Intelligence
Strategy. Indian Ocean Economies Study, (03), 132-
151+156.
IAMPA 2025 - The International Conference on Innovations in Applied Mathematics, Physics, and Astronomy
266