An Economic Solution to Copyright Challenges of Generative AI

For each phase, we start by discussing the potential impact of GenAI on consumer behavior and propose consumer-level research questions. The only exception is the testing stage in which, as we will discuss, GenAI can help firms partially replace customer inputs when conducting market research. These customer insights help us investigate how firms can harness the innovative potential of GenAI. In so doing, we bridge these emerging customer insights with corresponding firm marketing strategies, presenting research questions at the firm level (Hamilton, 2016).

“The Macroeconomics of Artificial Intelligence,” Brynjolfsson E, Unger G. International Monetary Fund, December 2023. As with most large systems, there were occasional outages when the system unexpectedly became unavailable. Workers who had previously been using the system now had to answer questions without access to it, and nonetheless they continued to outperform those who had never used the system. In the 1980s, expert systems, which consisted of hundreds or thousands of “if…then” rules drawn from interviews with human experts, helped diagnose diseases and make loan recommendations, but with limited commercial success.

Looking across major economies, a GenAI-driven productivity upswing could also make a substantial contribution to the global economy. We estimate that the lift to global GDP from stronger productivity could total $1.2t to $2.4t over the next decade. Enabled by data and technology, our services and solutions provide trust through assurance and help clients transform, grow and operate.

Generative AI and Its Economic Impact: What You Need to Know

For example, in India, where the number of individual creators is already on the rise, a survey of more than 1,600 freelancers found that 47% were using generative AI tools regularly and more than 50% reported a positive impact on their productivity. Meanwhile, as the Philippines strives to become Asia’s leading creative economy by 2030, generative AI can play a key role in professionalizing the work of the country’s freelancers. The magnitude of the productivity boost from GenAI will depend on the speed of its diffusion across organizations and industries. While GenAI has already spawned many innovations, it has yet to show a visible and meaningful boost in the aggregate productivity data. As we highlighted in our first article, the productivity boost from GenAI will likely occur with a lag as there has generally been a long delay between the inception of paradigm-shifting technologies and their diffusion across the economy and society.

We find that generative AI has the opposite pattern—it is likely to have the most incremental impact through automating some of the activities of more-educated workers (Exhibit 12). Pharma companies that have used this approach have reported high success rates in clinical trials for the top five indications recommended by a foundation model for a tested drug. This success has allowed these drugs to progress smoothly into Phase 3 trials, significantly accelerating the drug development process. Across the 63 use cases we analyzed, generative AI has the potential to generate $2.6 trillion to $4.4 trillion in value across industries. Its precise impact will depend on a variety of factors, such as the mix and importance of different functions, as well as the scale of an industry’s revenue (Exhibit 4). Treating computer languages as just another language opens new possibilities for software engineering.

Software engineers can use generative AI in pair programming and to do augmented coding and train LLMs to develop applications that generate code when given a natural-language prompt describing what that code should do. We estimate that generative AI could increase the productivity of the marketing function with a value between 5 and 15 percent of total marketing spending. Our updates examined use cases of generative AI—specifically, how generative AI techniques (primarily transformer-based neural networks) can be used to solve problems not well addressed by previous technologies. At present, training accounts for 80% of the energy usage and inference for about 20%, but, in the future, this is expected to flip on its head as the need for inference – passing new inputs through pre-trained models – accelerates. An often cited statistic, drawn from a paper by researchers at the Allen Institute for AI and the machine learning firm Hugging Face, is that generative AI systems can use up to 33 times more energy than machines running task-specific software. In the medium-to-long term, these concerns may have been alleviated by retrieval augmented generation (RAG).

In March 2023 alone, there were six major steps forward, including new customer relationship management solutions and support for the financial services industry. The era of generative AI is just beginning, and fully realizing the enormous benefits of the technology will take time. But business leaders should begin implementing generative AI use cases as soon as possible rather than waiting on the sidelines as the performance gap between laggards and early adopters will widen quickly. The competitive advantage will go to the organizations that are first to use generative AI to accelerate their business priorities, innovations, and company growth. As AI continues to grow in power, so too does the need for economic research to better understand how we can harness its benefits while mitigating its risks. Finally, innovations in AI systems may further improve the functioning of current AI tools.

A new report from McKinsey has put an estimate on these gains, predicting that generative technologies like ChatGPT, DALL-E, Google Bard, and DeepMind could add anywhere between $2.6 trillion to $4.4 trillion to the industry annually. While the use of AI has been simmering under the surface for decades, recent developments in generative AI have propelled the industry forward — opening up lucrative opportunities to countless businesses in its wake. Global economic growth was slower from 2012 to 2022 than in the two preceding decades.8Global economic prospects, World Bank, January 2023. Although the COVID-19 pandemic was a significant factor, long-term structural challenges—including declining birth rates and aging populations—are ongoing obstacles to growth.

As the field continues to evolve, we thought we’d take a step back and explain what we mean by generative AI, how we got here, and how these models work. The implication is that if we are looking for radical results from gen AI, then need to find big, difficult problems that gen AI can solve. Until then, we should continue experimenting with the technology to learn what it can do, but we should be careful about overcommitting resources without evidence of the feasibility, viability, and desirability of the innovation. Unlike the introduction of PCs, led by the killer app of the word processor, and the internet, driven by ecommerce, AI has no killer application to drive adoption. His research primarily focuses on digital transformation and innovation for global and multi-national organisations.

The expected business disruption from gen AI is significant, and respondents predict meaningful changes to their workforces. They anticipate workforce cuts in certain areas and large reskilling efforts to address shifting talent needs. Yet while the use of gen AI might spur the adoption of other AI tools, we see few meaningful increases in organizations’ adoption of these technologies. The percent of organizations adopting any AI tools has held steady since 2022, and adoption remains concentrated within a small number of business functions. Finally, this manuscript offers a roadmap to conduct research about the likely contributions of GenAI to marketing. GenAI often operates like a black-box and even its creators have limited understanding of its cognitive nature.

Second, previous papers outlining research questions about the effect of AI on marketing have focused on either the consequences on firms’ activity (Davenport et al., 2020; Huang & Rust, 2021, 2023) or consumer response (Puntoni et al., 2021). This article aims to bridge these two perspectives by developing research questions both at the consumer and firm level. In this way, we hope that this roadmap can help marketing scholars pursue research across various areas of specialization. The deployment of generative AI and other technologies could help accelerate productivity growth, partially compensating for declining employment growth and enabling overall economic growth. In some cases, workers will stay in the same occupations, but their mix of activities will shift; in others, workers will need to shift occupations.

Even if they don’t necessarily have to buy technological tools, they may need to train team members so they learn new skills. Some organizations have already utilized this process, offering 24/7 guidance and feedback to team members. Generative AI does skill-gap assessments and provides suggestions for learning courses and development ideas.

Its trustworthiness depends heavily on the quality of training datasets, according to the World Economic Forum. Generating content based on cumulative data input makes Gen AI worthwhile in many industries. The speed with which this technology can create content can help employees develop more content in less time and work more efficiently. This can reduce the need for human labor, raising concerns about job displacement and income inequality. The technology can also streamline class preparation and curriculum planning, enabling teachers to create personalized learning experiences based on an algorithmic analysis of student learning patterns and preferences. According to Access Partnership, this application of generative AI will lead to especially significant reprioritization of work activities for teachers in areas such as biological sciences, nursing, physics, geography, architecture, and computer science.

Ethics and Generative AI

This could empower teams to quickly access relevant information, enabling them to rapidly make better-informed decisions and develop effective strategies. Our estimates are based on the structure of the global economy in 2022 and do not consider the value generative AI could create if it produced entirely new product or service categories. The speed at which generative AI technology is developing isn’t making this task any easier. Until recently, a dominant trend in generative AI has been scale, with larger models trained on ever-growing datasets achieving better and better results. You can now estimate how powerful a new, larger model will be based on how previous models, whether larger in size or trained on more data, have scaled.

These large datasets come in many forms, such as text, images, audio recordings, and videos. Second, by being forced to predict parts of the inputs, self-supervised learning models develop a deeper understanding of the context. This enhanced understanding, coupled with training on significantly larger datasets, makes it easier to generalize to novel settings the economic potential of generative ai than was the case with past, supervised AI models. To fully unlock the potential of generative AI, life sciences companies must adopt a comprehensive approach and think in terms of connecting across value streams. This involves integrating AI across all workflows and building end-to-end processes and capabilities rather than focusing on isolated use cases.

As a result, one of the primary concerns is that they may lose their jobs, leading to social unrest. While Chat GPT is valid, its implementation may prove challenging for many companies. Professionals with remarkable technical expertise must be recruited so they can operate the algorithm effectively. Therefore, many organizations that can’t afford such additions may be left behind and make massive efforts to catch up to their competition. Marketing and advertising can already see the economic potential and gains of generative AI as they can create content based on their target audience’s preferences.

In factories where people operate complex machines and work with hazardous materials, avoiding accidents and ensuring safety are priorities. Machines and robots can perform these more laborious tasks with increased efficiency and without causing harm. As a result, companies don’t have to stress about extra costs resulting from job-related accidents, and employees can focus on other lower-risk tasks. We take a first look at where business value could accrue and the potential impacts on the workforce.

This is an aging network that is ill suited to respond to such sudden increases in demand. Great innovations often start out at a high cost, but as they reach a large market the costs to produce go down, so the price falls, enabling wider adoption. With generative AI use expected to grow rapidly this decade, there’s no time like the present to get these conversations going and processes put in place. Read the full report to discover potential use cases and opportunities, as well as what to consider if you’re thinking of using generative AI applications in your organization.

This licence allows anyone to reproduce OLJ articles at no cost and without further permission as long as they attribute the author and the journal. With more and more companies turning to LLMs for a competitive edge, training should be seen as “an ongoing expense,’ he adds. In the rush to invest in generative AI, one thing that may be overlooked is the actual costs involved in implementing it. AI has certainly closed the technology divide and developers of AI pair programmers may argue that in the long term, anyone could be a programmer. But these claims also deserve scrutiny, particularly claims that AI could replace human developers. Ever since the public got its hands on generative AI, and at periodic intervals throughout the release cycles of all the big developers’ major announcements, it’s been clear that generative AI output has a huge trust barrier to overcome.

A series of graphs show predicted compound annual growth rates from generative AI by 2040 in developed and emerging economies considering automation. This is based on the assumption that automated work hours are reintegrated in work at today’s productivity level. Two scenarios are shown for early and late adoption of automation, and each bar is broken into the effect of automation with and without generative AI. The addition of generative AI increases CAGR by 0.5 to 0.7 percentage points, on average, for early adopters, and 0.1 to 0.3 percentage points for late adopters.

We see a majority of respondents reporting AI-related revenue increases within each business function using AI. And looking ahead, more than two-thirds expect their organizations to increase their AI investment over the next three years. Our study focuses on art painting and logo design, two domains where copyright plays a pivotal role in safeguarding the integrity and commercial value of creative outputs. For art paintings, our research employs the WikiArt dataset [31], which comprises approximately 80,000 artworks spanning the last 400 years. This collection features pieces from over 1,000 renowned artists with a wide variety of styles and genres. For logo design, we use FlickrLogo-27 dataset [17], which consists of images from 27 distinct logo classes or brands, sourced from Flickr.

The technology’s improved ability to understand natural language has the potential to transform worker productivity by automating 60% to 70% of tasks that absorb employees’ time currently. The impact of generative AI is expected to be instrumental across all industries, especially in banking, high-tech, pharmaceuticals and medical products, and retail, McKinsey’s report says. The technology could add $200 billion to $340 billion in value to the banking industry, and $240 to $390 billion in value in retail. The potential benefits to the global economy from increased GenAI productivity could also be substantial. With the US market likely to remain at the forefront of GenAI investment, closely followed by Europe, Japan and China, global GDP could get a boost worth $1.2t (in our baseline scenario) and $2.4t (in the optimistic case) over the next decade. By executing and automating complex cognitive tasks that previously only humans could perform, GenAI has the potential to enhance workers’ efficiency, accelerate capital deepening and unlock substantial productivity gains across the economy.

To that end,

the company also recently announced the incorporation of generative AI capabilities into its human

resources software, Oracle Fusion Cloud Human Capital Management (HCM). Spending on the application of generative AI in the global banking industry is expected to increase from $386 million in 2023 to $8.499 billion in 2030, showing a significant growth trend. This growth was mainly reflected in the improvement of the efficiency and quality of service within the bank’s business. The advantage of generative AI technology is that it automatically generates text, performs data analysis, and provides accurate predictions.

After averaging 1.4% annually from 1973 to 1990, labor productivity growth accelerated to 2.2% between 1990 and 2000 and 2.7% between 2000 and 2007. Notably, TFP accounted for about half of the decade-long acceleration in labor productivity growth during the 1990s and early 2000s. In this installment, we explore the economic impact of GenAI through a productivity lens and quantify the extent to which the productivity potential of GenAI could bolster overall economic prospects in the next decade.

In the baseline condition, ChatGPT-4 received an identical prompt to that of the MBA students (i.e., generate ten ideas targeting college students in the U.S.). In the “prompted with good examples” condition, the authors supplemented the prompt with a set of highly rated ideas. They find that ideas generated by ChatGPT-4, regardless of the condition, exhibit higher average scores on purchase intentions compared to ideas generated by MBA students. Interestingly, they report no significant difference between the two GPT conditions (i.e., baseline versus prompted with highly rated ideas).

It’s likely to lead to increased efficiency and productivity but it may also lead to job displacement for some workers. A study by Accenture found that artificial intelligence could add $14 trillion to the global economy by 2035 with the most significant gains in China and North America. The study also predicted that AI could increase labor productivity by up to 40% in some industries.

Generative AI could have an impact on most business functions; however, a few stand out when measured by the technology’s impact as a share of functional cost (Exhibit 3). Our analysis of 16 business functions identified just four—customer operations, marketing and sales, software engineering, and research and development—that could account for approximately 75 percent of the total annual value from generative AI use cases. Our framework views the innovation process as a circular process that encompasses several interactions https://chat.openai.com/ with customers throughout the four stages. In the developing phase, firms frequently involve consumers in co-creation activities through open innovation platforms and crowdsourcing initiatives (Bayus, 2013; Cillo et al., 2021; Luo & Toubia, 2015; Rubera et al., 2016; Stephen et al., 2016). In the testing stage, firms conduct market research to gain customers’ perspective in order to select the one with the best chances of meeting market needs, given the various inputs generated in the previous phase (Kahn et al., 2006).

Of course, it’s possible that the risks and limitations of generative AI will derail this steamroller. Fine-tuning generative models to learn the nuances of what makes a business unique may prove too difficult,

running such computationally intensive models may prove too costly, and an inadvertent exposure of trade

secrets may scare companies away. A useful way to understand the importance of generative AI is to think of it as a calculator for open-ended,

creative content.

Individual roles will change, sometimes

significantly, so workers will need to learn new skills. Historically, however, big

technology changes, such as generative AI, have always added more (and higher-value) jobs to the economy

than they eliminate. Gathering and interpreting data is a crucial duty of HR professionals who need to identify patterns and predict employee behaviors. A big pharmaceutical company recently started using AI to process large sets of data and predict attrition rates in various departments.

Here is a small set of examples that demonstrate the technology’s broad

potential and rapid adoption. Given its potential to supercharge data analysis, generative AI is raising new ethical questions and

resurfacing older ones. Generative AI represents a broad category of applications based on an increasingly rich pool of neural

network variations. Although all generative AI fits the overall description in the How Does Generative AI

Work? Section, implementation techniques vary to support different media, such as images versus text, and to

incorporate advances from research and industry as they arise. Consider the challenges marketers face in obtaining actionable insights from the unstructured, inconsistent,

and disconnected data they often face.

But the limitations of these

early neural nets, combined with overhyped early expectations that could not be met due to those limitations

and the state of computational power at the time, led to a second AI winter in the 1990s and early 2000s. ChatGPT is the tool that became a viral sensation, but a multitude of generative AI tools are available for

each modality. In

image generation, Midjourney, Stable Diffusion, and Dall-E appear to be the most popular today. The latest Insights article from our member Taylor Wessing highlights the key changes to the Commission’s draft Artificial Intelligence Act, or AI Act.

Therefore, a significant amount of human supervision is required for conceptual and strategic thinking tailored to the needs of the specific organization. We hope this research has contributed to a better understanding of generative AI’s capacity to add value to company operations and fuel economic growth and prosperity as well as its potential to dramatically transform how we work and our purpose in society. Companies, policy makers, consumers, and citizens can work together to ensure that generative AI delivers on its promise to create significant value while limiting its potential to upset lives and livelihoods.

This reduces the time staff must spend

collecting demographic and buying behavior data and gives them more time to analyze results and brainstorm

new ideas. Generative AI can significantly change the face of business operations, as employees might spend less time on manual tasks that algorithms automate and streamline. Therefore, professionals may spend more time on human interactions, communicating more efficiently with each other instead of being buried under large sets of data. Chatbots could even become people’s companions as they guide them through daily activities and act as their personal assistants. However, companies should be aware of the potential pitfalls involved before implementing AI in their organizations.

As generative AI creates content based on existing material, doesn’t that mean that it infringes upon copyrights? This consideration creates the necessity for new regulations and legal frameworks to ensure algorithms are used ethically. Professionals must also refrain from copying content verbatim since they may receive copyright strikes.

Marketing functions could shift resources to producing higher-quality content for owned channels, potentially reducing spending on external channels and agencies. Generative AI refers to deep-learning models that can take raw data — say, all of Wikipedia or the collected works of Rembrandt — and “learn” to generate statistically probable outputs when prompted. At a high level, generative models encode a simplified representation of their training data and draw from it to create a new work that’s similar, but not identical, to the original data. The applications for this technology are growing every day, and we’re just starting to explore the possibilities. At IBM Research, we’re working to help our customers use generative models to write high-quality software code faster, discover new molecules, and train trustworthy conversational chatbots grounded on enterprise data.

Compared to earlier forms of AI and analytics, such as machine learning and deep learning, generative AI could increase productivity by up to 40 percent. Adoption is also likely to be faster in developed countries, where wages are higher and thus the economic feasibility of adopting automation occurs earlier. Even if the potential for technology to automate a particular work activity is high, the costs required to do so have to be compared with the cost of human wages. In countries such as China, India, and Mexico, where wage rates are lower, automation adoption is modeled to arrive more slowly than in higher-wage countries (Exhibit 9).

They could also have an impact on knowledge workers whose activities were not expected to shift as a result of these technologies until later in the future (see sidebar “About the research”). For one thing, mathematical models trained on publicly available data without sufficient safeguards against plagiarism, copyright violations, and branding recognition risks infringing on intellectual property rights. A virtual try-on application may produce biased representations of certain demographics because of limited or biased training data.

There will be a multitude of additional innovation that is not included in many of these forecasts. For example, there are storage technologies that can save 30% of the cost of energy for running data centers if they become widely deployed. However, even with these facts that are not taken into account, the chances that we can support the growth of data centers needed to support AI at scale is questionable.

A foundation model is a large, pre-trained model used as a base for developing more specialized and task-specific models (Bommasani et al., 2021). Specifically, they create new content (e.g., text, image, video, data) by using patterns learned during training to predict the next item in a sequence. For instance, OpenAI and Microsoft have deployed GPT-3 in a variety of downstream tasks, such as Bing, Duolingo, GitHub Co-pilot, and ChatGPT. To understand how foundation models produce new content, let us take the example of Large Language Models (LLMs), a subset of foundation models that have gained significant prominence as they are trained to facilitate user interaction through natural language. A language model (LM) is a statistical representation of a language, which computes the probability of a given sequence (a word, phrase, or sentence) occurring in this language. Similar to LMs, LLMs are trained in a self-supervised logic to predict a masked word within a sequence of words.

This would largely impact high earners like knowledge workers and could add “trillions of dollars in value to the global economy,” McKinsey said. In these major domains, GenAI stands not just as a tool but as a transformative force, reshaping the way tasks are approached and executed, which can lead to unprecedented levels of efficiency and innovation. In the next installment of this series, we will examine the labor-augmenting capabilities of GenAI across sectors and occupations in greater detail. Using the ICT period as a reference, we created three scenarios – trend, revival (our baseline) and boom – that correspond to three different productivity outcomes for the next decade. Our analysis builds on the scenarios developed in the previous chapter on capital investment.

Generative AI (GAI) is the name given to a subset of AI machine learning technologies that have recently. developed the ability to rapidly create content in response to text prompts, which can range from short and. simple to very long and complex. Different generative AI tools can produce new audio, image, and video. content, but it is text-oriented conversational AI that has fired imaginations. In effect, people can. You can foun additiona information about ai customer service and artificial intelligence and NLP. converse with, and learn from, text-trained generative AI models in pretty much the same way they do with. humans. So, along with its remarkable productivity prospects,. generative AI brings new potential business risks—such as inaccuracy, privacy violations, and intellectual. property exposure—as well as the capacity for large-scale economic and societal disruption. For example,. generative AI’s productivity benefits are unlikely to be realized without substantial worker retraining. efforts and, even so, will undoubtedly dislocate many from their current jobs. Consequently, government. policymakers around the world, and even some technology industry executives, are advocating for rapid. adoption of AI regulations.

But a full realization of the technology’s benefits will take time, and leaders in business and society still have considerable challenges to address. These include managing the risks inherent in generative AI, determining what new skills and capabilities the workforce will need, and rethinking core business processes such as retraining and developing new skills. The pace of workforce transformation is likely to accelerate, given increases in the potential for technical automation.

On the flip side, there’s a continued interest in the emergent capabilities that arise when a model reaches a certain size. Encoder-decoder models, like Google’s Text-to-Text Transfer Transformer, or T5, combine features of both BERT and GPT-style models. They can do many of the generative tasks that decoder-only models can, but their compact size makes them faster and cheaper to tune and serve. Decoder-only models like the GPT family of models are trained to predict the next word without an encoded representation.

We’re even using generative AI to create synthetic data to build more robust and trustworthy AI models and to stand-in for real data protected by privacy and copyright laws. Our latest survey results show changes in the roles that organizations are filling to support their AI ambitions. In the past year, organizations using AI most often hired data engineers, machine learning engineers, and Al data scientists—all roles that respondents commonly reported hiring in the previous survey.

For instance, generative AI’s ability to identify leads and follow-up capabilities could uncover new leads and facilitate more effective outreach that would bring in additional revenue. Also, the time saved by sales representatives due to generative AI’s capabilities could be invested in higher-quality customer interactions, resulting in increased sales success. Following are four examples of how generative AI could produce operational benefits in a handful of use cases across the business functions that could deliver a majority of the potential value we identified in our analysis of 63 generative AI use cases. In the first two examples, it serves as a virtual expert, while in the following two, it lends a hand as a virtual collaborator. We then estimated the potential annual value of these generative AI use cases if they were adopted across the entire economy. For use cases aimed at increasing revenue, such as some of those in sales and marketing, we estimated the economy-wide value generative AI could deliver by increasing the productivity of sales and marketing expenditures.

• Generative AI technology is built on neural network software architectures that mimic the way the human

  brain is believed to work.

• This is based on the assumption that automated work hours are reintegrated in work at today’s productivity level.
• On the one hand, GenAI demonstrates superior capabilities in generating empathy and unique content that is specifically tailored to consumer needs, compared to other types of AI (Huang & Rust, 2023).
• This article aims to bridge these two perspectives by developing research questions both at the consumer and firm level.
• The modeled scenarios create a time range for the potential pace of automating current work activities.
• This style of training results in an AI system that can output what humans deem as high-quality conversational text.

However, these modifications may compromise model performance due to either the exclusion of high-quality, copyrighted training data from training or restrictions on content generation [19]. The complexity and ambiguity of copyright law add another layer of difficulty, blurring the line between infringing and non-infringing outputs. The resulting uncertainty could lead to a significant waste of resources on both sides while these issues are debated in courts [32]. The marketing field has recognized that new technologies often alter consumer behavior (Giebelhausen et al., 2014; Hoffman & Novak, 2018). Accordingly, we argue that research at the intersection of marketing, neuroscience, and cognition should investigate how GenAI will affect the cognitive capabilities of consumers, especially with respect to creativity.

Because they are trained on large amounts of data as well as frequently trained on different languages, they are called Large LMs. If we consider words and punctuation signs as tokens, we can depict an LLM as a conditional probability distribution p(xn|x1, …, xn − 1) over tokens, in which each xi is drawn from a fixed vocabulary. An LLM generates text by iteratively sampling from the learned distribution to select the next token.

However, it will take time and human expertise to unlock their full potential in a way that’s responsible, trustworthy and safe. If you’re considering using generative AI applications, it’s important to establish a set of internal processes and controls for everyone in your organization to follow. Microsoft, for example, teams up with Caterpillar in combining AI technology with industry expertise to optimise crop management. Health tech startup Alken, in partnership with pharmaceutical companies, uses AI to streamline clinical trials while protecting proprietary data .

Walmart uses predictive analytics and automation to more accurately forecast demand and reduce excess inventory and shortages, thereby improving the efficiency of the overall supply chain. In supplementary materials Section A.2, we explore the SRS when the AI developer is considered a special player within a game characterized by a permission structure. Specifically, Figure 6 (a) shows the result of SRS for an AI-generated painting in Van Gogh’s style, and Figure 6 (b) shows the result of SRS for an AI-generated logo for Sprite. Both figures show that the AI developer achieves a markedly higher SRS compared to training data contributors.