Application and Impact of Artificial Intelligence in Digital

Marketing: Take TikTok as an Example

Manfei DENG GUO

School of Business, Guangdong University of Foreign Studies, Guangzhou, 511400, China

Keywords: Artificial Intelligence (AI), Digital Marketing, TikTok, Algorithmic Recommendation System, Data Privacy.

Abstract: Nowadays, artificial intelligence is widely used in digital marketing. TikTok has a unique algorithmic

recommendation system and large user base, making it a crucial case for studying AI's role in digital marketing.

Throughout a case studies, this study explores how AI is applied in TikTok's digital marketing, aiming to

understand its application effectiveness, limitations, and risks. The research finds that AI significantly

improves TikTok's marketing effectiveness. Artificial intelligence enables precise targeting via personalized

recommendations, offers new marketing ways through virtual influencers, enhances real - time interactions

in live - streaming, and provides valuable insights through data analysis. However, there are also many

drawbacks. Data privacy issues, like over - collection and misuse, and algorithmic bias, which limits content

diversity. To address these problems, TikTok can implement a comprehensive privacy policy, incorporate

diversity metrics into algorithms, and offer users more control over content recommendations. These

measures can enhance user experience and platform sustainability.

1 INTRODUCTION

With the rapid development of digital technology,

artificial intelligence (AI) is increasingly being used

across a wide range of industries, especially in digital

marketing. As one of the most popular social media

platforms in the world, TikTok has become a new

bridge for brands and customers to interact, with its

unique algorithmic recommendation system and large

user base. Besides, AI technologies play a crucial role

in enhancing marketing campaigns by leveraging

advanced algorithms to analyze vast amounts of user

data, enabling personalized content

recommendations, efficient ad targeting, and real-

time performance optimization.

This study is dedicated to evaluate the application

effectiveness of AI on TikTok and has two purposes.

The first is to gain a comprehensive understanding of

the use of AI in TikTok digital marketing. The second

is to identify the limitations and risks associated with

these applications. Through case studies, this essay

will explore how AI is utilized for personalized

recommendations, virtual influencer applications,

live e-commerce, interactive enhancements, data

analysis and trend prediction on TikTok.

Furthermore, this study will identify limitations and

risks of AI marketing. The existing data privacy

problems, algorithm deviations and other problems

are analyzed, and corresponding improvement

suggestions are put forward. Also, by analyzing

current applications and their effectiveness, people

can gain a deeper understanding of how brands are

leveraging these technologies to achieve their

marketing goals. In addition, exploring the unique

ecosystem of TikTok can help address emerging

challenges and pave the way for innovative solutions

in digital marketing. While the existing literature

extensively addresses the role of AI in digital

marketing, but there are clear gaps in the detailed

analysis of TikTok. For instance, one study highlights

that AI significantly enhances customer engagement

through personalized content and efficient ad

targeting (Labudová, 2024), dresses the role of AI in

digital marketing, but there are clear gaps in the

detailed analysis of TikTok. Another relevant work

emphasizes the importance of AI in analyzing user

behavior and preferences (Duan, 2024). However,

such studies often do not delve deeply into TikTok's

distinct features and user dynamics.

The structure of this essay is follow: Firstly, this

essay will discusses the current state of AI in digital

marketing, with a focus on accurate user analytic,

precise ad placement, and content generation and

682

DENG GUO, M.

Application and Impact of Artiﬁcial Intelligence in Digital Marketing: Take TikTok as an Example.

DOI: 10.5220/0014001400004916

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 2nd International Conference on Public Relations and Media Communication (PRMC 2025), pages 682-687

ISBN: 978-989-758-778-8

optimization. Secondly, a case study about TikTok

and a statement about the role of artificial intelligence

in personalized recommendation system, virtual

influencers, live streaming, and data analytics will be

presented. Furthermore, this essay will explores

possible risks, such as data privacy and algorithmic

bias, and suggests improvements. Finally, this essay

will summarizes key findings and outlines directions

for future research.

2 THE APPLICATION STATUS OF

ARTIFICIAL INTELLIGENCE

IN DIGITAL MARKETING

2.1 Accurate User Profile Construction

and Personalized

Recommendations

In the era of digital marketing, accurate user analytic

and personalized content recommendations are

crucial for brands to achieve more precise targeting

and improve advertising effectiveness. As the Internet

continues to flourish, it has accumulated massive and

diverse information data. These data not only cover

users' past shopping records and detailed payment

records, but also include precise information such as

recipient addresses (Zhou & Zhao, 2024).With

advanced algorithms and in-depth analysis

capabilities, AI technology is able to dig deeper into

the huge value latent in various data and information.

Using artificial intelligence (AI), brands can analyze

large data sets such as browsing history, purchasing

behavior, interaction patterns, demographic

information, and sentiment analysis, as machine

learning algorithms identify patterns and correlations

in these data points and categorize them into different

categories based on user behavior and interests. This

process helps brands pinpoint the core needs of their

users and provides extremely important support for

companies planning marketing campaigns. In terms

of personalized recommendations, AI uses content-

based collaborative filtering techniques to customize

content and product recommendations. As a result,

brands can accurately deliver personalized ads to

specific user groups, optimizing user stickiness and

conversion rates. In this way, enterprises can, on the

one hand, effectively control marketing costs and

avoid unnecessary waste of resources; on the other

hand, they can also significantly improve the

precision and relevance of their marketing

campaigns, thereby dramatically increasing

marketing effectiveness and realizing the efficient use

of resources and maximization of business value

(Zhou & Zhao, 2024).

2.2 Precision Advertising Delivery

Artificial Intelligence and Big Data technologies are

able to process massive amounts of data, optimize

marketing decisions, reduce labor costs, and improve

the accuracy of advertising, thus reducing marketing

costs (Du, 2024). For example, AI tools can monitor

ad effectiveness in real time and dynamically adjust

ad strategies by selecting the best release times and

delivery channels that match the target audience's

activity times and interests. In addition, AI can also

analyze key metrics such as ad serving size, click-

through rate and conversion rate to detect ad serving

problems early. If it is found that ads are ineffective

during certain times of the day or on certain channels,

AI will quickly and dynamically adjust ad serving

parameters, such as ad content, ad serving frequency

and ad budget allocation, based on predefined or

machine-learning optimization strategies to ensure

the best ad serving results. By delivering ads to users

during their active periods and within their interests,

AI increases the likelihood that users will engage with

the ads. This will results in higher click-through rates,

indicating greater user interest and engagement.

2.3 Content Generation and

Optimization

AI also plays an important role in the content

marketing space, with its ability to automatically

create content and turn data into readable narratives

through its natural language generation technology.

This capability allows for the rapid production of

personalized, relevant content, improving

engagement and efficiency. In video production, tools

like CapCut and the AI-driven automatic editing

function in Adobe Premiere Pro can analyze video

clips and suggest edits, effectively saving time and

improving the overall quality. When it comes to

image generation, platforms such as Midjourney are

utilized, which held a 9.90% market share as of July

2023 (Labudová, 2024).

In the realm of written content, ChatGPT has

emerged as a dominant tool, boasting a 75.20% usage

rate among marketing and advertising professionals.

Moreover, according to Labudová's (2024) research,

52.5% of professionals use AI for writing content and

creative materials. Clearly, AI significantly enhances

the efficiency and creativity in different aspects of

content creation.

Application and Impact of Artiﬁcial Intelligence in Digital Marketing: Take TikTok as an Example

683

3 CASE STUDY: ARTIFICIAL

INTELLIGENCE FOR DIGITAL

MARKETING AT TIKTOK

3.1 Application of Personalized

Recommendation System

Artificial Intelligence on TikTok influences users

through personalized content recommendations to

improve user experience and increase engagement.

Furthermore, TikTok uses AI algorithms and large

datasets to analyze user behavioral data such as likes,

comments, and viewing times, as well as

demographic information such as age and location.

This provides detailed user profiles for personalized

recommendations, helping brands and creators to

accurately identify their target audience so they can

effectively optimize their content and advertising.

Besides, AI algorithms can not only provide

personalized content based on users' interests and

preferences, but also discover new content relevant to

their interests, thereby improving TikTok's overall

user experience. A large number of respondents

appreciated TikTok's personalized content

recommendation feature, which helps them discover

new content that matches their tastes and interests and

helps them stay engaged and entertained (Loke,

2023).

3.2 Virtual Influencer Applications

TikTok uses artificial intelligence to create virtual

influencers that autonomously generate personalized

images of AI influencers based on advertisers' scripts

and marketers' needs, allowing virtual celebrities to

connect with real celebrities and participate in short

video sessions. This trend can increase interest in the

content and make it more accessible to a wider

audience. Furthermore, Virtual influencers can not

only promote products without rest, but also eliminate

the risk of negative news that real influencers may

have, thus reducing marketing costs.

Meanwhile ， virtual influencers have realistic

characteristics, features and traits of humans.

According to a 2022 study by The Influencer

Marketing Factory, at least 58% of respondents

follow at least one virtual influencer, and 35% have

purchased a product promoted by a virtual influencer

(Bringé, 2022). In addition, the number of virtual

influencers in 2015 was nine, and by 2022, the

number has exceeded 200. Virtual influencers offer

greater control and lower expenses, but they lack

authenticity and can have a negative impact on the

younger demographic (Cowan, 2022). Nevertheless,

virtual influencers have a high level of engagement

among young people, and their market value reached

$4.6 billion in 2022 and is expected to grow 26% by

2025. Brands such as Puma, Alibaba, and Samsung

are already using virtual influencers to increase

engagement with younger groups (Plazibat &

Marunica, 2024).

3.3 AI Live Streaming and Interactive

Enhancement

The trend of AI live streaming is on the rise in

TikTok. AI-enabled web hosts are becoming an

innovative feature. For instance, certain beauty

brands on TikTok have deployed AI hosts to host

makeup tutorials. These AI hosts, equipped with

natural language processing capabilities, can

smoothly follow pre-programmed scripts. Also, the

AI is involved in real-time data-driven interactions. If

a large number of viewers comment on a specific

eyeshadow color, the AI anchor can immediately

provide more relevant color demonstrations and how

to apply them, thus providing viewers with a novel

and engaging visual experience. Besides, in the

context of TikTok's live - streaming, AI can promptly

filter and review incoming barrages, effectively

shielding illegal ones. By continuously updating its

filtering mechanisms, AI ensures that users can enjoy

a better experience and contributes to shaping a

healthier cyberspace (Jiang, 2024).

What is more, AI also plays a crucial role in the

TikTok live e-commerce space. Artificial

intelligence-based smart customer service can

instantly answer users' frequently asked questions.

Taking the popular live streaming of electronic

products as an example, AI-based customer service

can quickly provide accurate data when users ask

about the battery life of their smartphones.

Additionally, an auto-respond system makes another

AI-based feature that ensures user queries are not

ignored during a live stream. For example, in a live

clothing broadcast, if a user asks if a specific size of

a dress is available, the auto-respond system can

check inventory and respond in a timely manner,

enhancing the user experience and promoting higher

conversion rates during e-commerce.

Nowadays, numerous companies and news

publishers are now piloting and implementing

artificial intelligence (AI) technology within their

management platforms. Instead of relying on human

moderators to sift through individual comments, they

are leveraging AI - driven technology. This is because

AI has the capacity to efficiently handle a vast

PRMC 2025 - International Conference on Public Relations and Media Communication

684

number of platform comments, streamlining the

moderation process and enhancing overall platform

management (Jiang, 2024).

3.4 Data Analysis and Trend Prediction

In TikTok, data analytics are key to enhancing live AI

broadcasts and interactions. By comprehensively

analyzing user behavior data, content data, and social

interaction data, TikTok mines valuable insights and

gives creators the resources they need to thrive in the

dynamic world of live content creation, while

optimizing advertising strategies for maximum

effectiveness.

In terms of user behavioral data, the platform

tracks behaviors such as video viewing, liking,

commenting and sharing. Patterns in these behaviors

can reveal the genres of content that users are most

interested in at any given time. For example, if there

is a sudden spike in views and interactions on a dance-

related video, TikTok can quickly identify dance as a

trending topic. In this way, the platform can highlight

currently popular challenges and topics to drive more

user engagement.

Content data, including video tags, descriptions,

and content categories, is equally important.

Analyzing this data helps TikTok understand what

types of content resonate with its audience. Creators

have access to a range of tools and tips. Meanwhile,

they can access detailed analytic reports on video

performance, such as viewer retention at different

time intervals in a video. This information can help

creators adjust their content strategy. In addition,

TikTok uses data to recommend creative inspiration

and popular music to creators. If a certain music track

has been used in a number of highly engaging videos,

that music can be recommended to other creators to

increase the appeal of their new content.

Social interaction data such as followers, follower

lists, and group interactions can also be utilized.

These data helps TikTok understand the influence of

social networks in the platform. Based on these

insights, the platform can push currently trending

content or upcoming events with targeted ads. For

example, if a large group of users in a particular

network are interested in a fitness challenge, ads

promoting a new fitness-related live event can be

pinpointed to them to maximize the reach and impact

of the event.

4 POTENTIAL RISKS OF

CURRENT APPLICATIONS

AND SUGGESTIONS

4.1 Data Privacy Issues

In order to provide more accurate user analytics and

content recommendations through AI algorithms,

TikTok can collect a large amount of data. These data

will includes many factors such as direct marketing

data as well as user location and device information.

While the platform claims that this data is used to

improve the user experience, there is a high risk of

over-collection and possible misuse of data.

Besides that, AI algorithms also collect and

aggregate data paths indirectly through tracking

technologies, such as browsing history, time on page,

and user interactions. Users are often passive in this

process, unaware of what data is being collected and

how it is being used, which highlights the hidden and

opaque nature of data collection.

What is more, storage security also poses

significant risks. Large amounts of user data are

stored on servers. A security breach in the storage

system could lead to a data breach. For example,

hacking or internal employee misuse could

compromise stored user data, putting personal

information at risk (Wang, 2023).

4.2 Algorithmic Bias

TikTok's recommendation algorithm is based on

historical user behavior data, which may bias

recommendations. For example, if a user initially

watches a large number of videos of a certain type or

topic (e.g., pet videos) due to random factors, the

algorithm may incorrectly assume that the user is only

interested in that type of content. As a result, the

algorithm will continue to recommend similar videos

and ignore other content that the user may be

interested in. Based on the user's past behavior, they

have access to only a narrow range of information,

thus limiting their exposure to a broader and more

diverse range of content, creating a homogenized

content ecosystem, and reducing the diversity and

vibrancy of the community as a whole. Over time,

these biases degrade the user experience and reduce

platform engagement. Recognizing and

understanding these algorithmic biases is critical to

improving recommendation systems, increasing user

satisfaction, and maintaining a healthy and vibrant

platform ecosystem.

Application and Impact of Artiﬁcial Intelligence in Digital Marketing: Take TikTok as an Example

685

4.3 Suggestion

To address the issues of algorithmic bias and enhance

user experience on TikTok, several measures can be

implemented. Firstly, TikTok should develop a

comprehensive data protection mechanism that

standardizes data management processes, covering

data collection, storage, usage, and transmission,

because ensuring users' rights to be informed, choose,

access, and delete their data is crucial. Besides,

TikTok should set up clear penalties for illegal data

acquisition and abuse, so as to effectively protect

users' privacy and data rights.

Secondly, TikTok should incorporate diversity

assessment metrics into the recommendation

algorithm beyond just matching user historical

behavior. For example, Diversity scores based on

content categories, topics, and styles can be

calculated by TIKTOK. This can ensures that

recommendations maintain relevance while covering

a broader range of topics, preventing over-

concentration on similar content types. By doing this,

the platform can offer more varied and engaging

content to its users.

Thirdly, for new users or those with limited

behavioral data, Tiktok should use a restart

recommendation strategy based on content features

rather than relying solely on initial interactions.

Besides, general video characteristics such as hot

topics and popular elements can be leveraged by

tiktok to provide users with diversified initial

recommendations.This approach helps users discover

a wider array of potentially interesting content,

broadening their interest boundaries from the outset.

Moreover, TikTok can offer clear interest

adjustment tools within the app interface, such as an

"Interest Preferences" button. Users can actively add

or remove content categories they are interested in,

adjust the weight of existing interest tags, and directly

communicate their desire for more diversified content

to the algorithm. This feature empowers users to

customize their content feed according to their

evolving preferences, enhancing personalization and

engagement. By implementing these suggestions,

TikTok can mitigate algorithmic biases, enhance user

satisfaction, and promote a healthier and more

dynamic platform ecosystem. These measures not

only improve the quality and variety of content

recommendations but also ensure greater

transparency and control for users over their data and

preferences.

5 CONCLUSION

This study examines the use and impact of Artificial

Intelligence (AI) in digital marketing, with a focus on

TikTok. Through a detailed case study, this essay

analyze how TikTok uses AI for personalized

recommendations, virtual influencer adoption, real-

time e-commerce, interactive enhancement, data

analytics, and trend forecasting. The results show that

AI greatly improves marketing effectiveness in

TikTok by enabling more precise targeting,

increasing user engagement, and facilitating the

creation of innovative content. Artificial intelligence-

driven personalized recommendation systems

leverage large amounts of user data to deliver

personalized content and ads that maximize user

engagement and conversion rates. The ability of

online virtual celebrities to continuously promote

without interruption provides brands with a new way

to engage young people. In addition, AI-powered live

streaming and customer service tools can enhance

real-time interactions and provide users with accurate

and timely information, thereby increasing user

satisfaction and conversion rates. By analyzing data

on user behavior, content, and social interactions,

TikTok can identify trending topics and issues,

providing creators with valuable insights and

resources to help them succeed. These data-driven

strategies not only enrich the user experience, but also

help creators improve their content strategy,

contributing to the overall success of the platform.

However, there are also some drawbacks to the

current use of AI on TikTok. Data privacy issues are

of great public concern, including problems such as

excessive collection and potential misuse of user data.

Users often remain unaware of the extent of data

collection and usage, highlighting the opaque nature

of these practices. Meanwhile, algorithmic bias poses

a significant risk, where historical user behavior data

may limit exposure to diverse content,which degrades

the user experience. These risks pose a challenge to

the sustainability of the platform.

To mitigate these risks, TikTok can employ a

variety of strategies. These risks can be mitigated by

implementing a comprehensive privacy policy,

incorporating diversity metrics into its algorithms,

providing content recommendations and features for

new users, and offering interest matching tools.

Through these measures, TikTok can ensure that

users are given the right to know, the right to choose,

the right to access and the right to delete data. At the

same time, illegal access and misuse of data is greatly

reduced to effectively protect user privacy. In

addition, it avoids over-concentration of users on

PRMC 2025 - International Conference on Public Relations and Media Communication

686

similar content types, thus enhancing the diversity

and richness of the platform's content ecosystem, and

allowing new or data-scarce users to discover more

diverse content. This also allows users to add or

remove content categories, adjust existing interest

tags, and express their desire for more diverse content

directly to the algorithm.

REFERENCES

Bringé, A. 2022. The rise of virtual influencers and what it

means for brands: Forbes 2022: Retrieved 18. October

2023, from https://www.forbes.com/sites/forbe

scommunicationscouncil/2022/10/18/the-rise-of-

virtual-influencers-and-what-it-means-for-brands/

Cowan, P. 2022. The pros and cons of virtual influencers:

Smartosc 2023: Retrieved 22. April 2023, from

https://www.smartosc.com/insights/the-pros-and-cons-

of-virtual-influencers

Du, L. G. 2024. Analysis of the application of artificial

intelligence and big data technology in digital

marketing. Marketing World (10):5-7.

Duan, J. 2024. Identification and influence of tourism

consumption behavior based on artificial

intelligence. Informatica 48(15):135-150.

Jiang, J. 2024. Availability of AI and AI anchors in social

media—Take TikTok as an example. Lecture Notes in

Education Psychology and Public Media 58:291-297.

Labudová, L. 2024. The influence of artificial intelligence

on modern marketing. Academic Conferences

International Limited.

Loke, L. 2023. The impact of AI integration on audience: A

qualitative study of young adults' perspectives and

attitudes towards the integration of AI on TikTok.

Faculty of Humanities and Social Sciences, Media and

Communication, Digital Media and Analysis, C-thesis

15 HP, Karlstad University.

Plazibat, I., & Marunica, S. 2024. The impact of artificial

intelligence and of The #deinfluencing trend on

influencer marketing. Varazdin: Varazdin

Development and Entrepreneurship Agency (VADEA).

Wang H. 2023. Application of artificial intelligence and big

data technology in digital marketing. Information and

Computer (Theoretical Edition) (12): 29-35.

Zhou, Y. & Zhao, C. 2024. Analysis of innovative

applications of artificial intelligence and big data

technologies in digital marketing. Market Weekly

37(21): 100-103.

Application and Impact of Artiﬁcial Intelligence in Digital Marketing: Take TikTok as an Example

687