**2. Literature review**

#### **2.1 Algorithmic bias and its implications for YouTube**

Algorithmic bias is a phenomenon that occurs when machine learning algorithms produce results that are systematically skewed toward certain groups or types of content [26]. In the context of online platforms, such as YouTube, algorithmic bias can have significant implications for content creators and consumers. According to [27, 28], the platform's recommendation algorithm may steer users toward certain types of content based on their previous viewing history. To some extent, this may lead to a homogenization of content and limit the exposure of lesser-known creators [29, 30]. Additionally, YouTube's monetization policies and algorithms may favor certain types of content, such as those that are advertiser-friendly, leaving other talented creators with limited monetization options [31–33].

As algorithmic bias has the potential to negatively impact content creators from all backgrounds, it is particularly problematic for filmmakers from developing countries. These creators often face significant challenges in accessing the resources necessary to produce high-quality content, and their work may not fit within the dominant algorithmic frameworks [34–37]. While some scholars argue that algorithmic bias may have a positive impact on the quality of content by promoting popular, highquality works [38, 39], others suggest that it perpetuates existing power imbalances and limits diversity and representation, especially with regard to cultural production and marginalization [40–42].

#### **2.2 Monetization potential and algorithmic bias in film content**

Concerns have been raised about the impact of geo-restrictions on the monetization potential of film content on YouTube, particularly for creators from developing countries [43]. As a result, their content may be less likely to appear in search results or be recommended to viewers across different geographical regions, reducing their visibility and limiting their ability to monetize their work, ultimately, reducing its potential audience and limiting their ability to generate revenue [44, 45]. This is because, the algorithms that determine what content is recommended or promoted on platforms like YouTube are often based on metrics such as views, likes, and audience engagement [46], which can disadvantage content that has not yet gained a significant following.

Additionally, algorithmic bias may also lead to content creators being categorized and pigeonholed, with their work being assigned labels that limit its reach or relevance. To some extent, some viewers may flag specific content, deeming it insensitive to particular group or societal issues [47, 48]. This is the case for most of the marginalized communities or least developed countries. For example, research

has shown that algorithms can perpetuate and amplify stereotypes and discrimination, especially toward historically marginalized groups. Consequently, [49] argues that thinking of ethical practice in algorithmic predictions and digital discrimination requires a fundamental reconsideration of justice, fairness, and ethics, citing that, it goes beyond mere methodologies and tools, and emphasizes the need for a habit of ethical thinking in the wake of machine learning. Therefore, by integrating relational ethics into data science practices, a more comprehensive and responsible approach can be achieved, promoting fairness and justice in algorithmic decision-making [50].

Despite the potential negative impact of algorithmic bias and geo-restrictions on the monetization potential of film content on YouTube, there is a lack of research on the specific challenges faced by filmmakers in developing countries like Botswana. This study, therefore, aims to address this gap by examining the limitations faced by Botswana film content creators on YouTube, with specific attention toward algorithmic bias and geo-restrictions, and their impact on the monetization potential for this content. To achieve this goal, the study answers the following research questions:

