AI and the Issue of Human-centricity In Copyright Law

Syllabus: GS 3 – Science and Technology- Developments and their Applications and Effects in Everyday Life. Awareness in the field of IT, and issues relating to Intellectual Property Rights.

Context: Across the globe there are significant change in attitude of global leaders towards AI regulation, including in India.

Source: The Hindu | Editorial dated 02 November 2023

AI and the 21st Century

• Artificial intelligence is the emulation of human intelligence processes using machines, particularly computer systems.
• AI is a transformative technology of the 21st century, reshaping various sectors and the job market.
• It has evolved rapidly, leading to significant changes in how we live and work.
• AI research dates back to the 1950s and 1960s, with early efforts to mimic human thinking.
• Advances in AI have enabled tasks like speech recognition, computer vision, and machine learning.
• In 2010, Apple introduced Siri, an intelligent personal assistant, enhancing human-computer interaction.
• Amazon’s Alexa further popularized AI-powered personal assistants, becoming a global phenomenon.
• AI has become a cornerstone for leading and prosperous companies today, including Alphabet, Apple, Microsoft, and Meta.
• These companies employ AI technologies to enhance their operations and stay ahead of competitors.
• Investment in autonomous technology by car manufacturers hints at a future with self-driving cars.
• AI’s future involves enhanced communication with humans through machine learning and recognition.
• Technologies like Google Duplex demonstrate AI’s ability to converse like humans and perform tasks.

Types of Artificial Intelligence:

Based on Capabilities:

• Narrow AI (Weak AI): Focuses on specific, limited tasks and lacks broader cognitive abilities.
• Examples include Apple Siri, IBM Watson, and Google Translate.
• General AI (Strong AI): Possesses the capability to understand and learn any intellectual task like a human but remains a challenging goal.
• Fujitsu’s K computer and Microsoft’s investment in OpenAI are notable attempts.
• Super AI: Hypothetical AI surpassing human intelligence and emotions, capable of independent thinking and decision-making.

Based on Functionalities:

• Reactive Machines: Operate solely based on current data and do not store past experiences.
• IBM’s Deep Blue, which defeated chess grandmaster Garry Kasparov, is an example.
• Limited Memory: Utilize past data for decisions within a specific time frame, commonly seen in self-driving vehicles.
• Theory of Mind: An advanced concept AI, yet to be fully realized, understands human emotions, thoughts, and behavioral changes.
• Examples include Kismet and Sophia.
• Self-awareness: A hypothetical AI that understands its own internal state and human emotions, potentially smarter than humans.

Branches of AI:

• Machine Learning: Develops algorithms that learn from data, used in applications like image recognition and spam filtering.
• Deep Learning: Utilizes artificial neural networks to gain knowledge from data, solving problems such as NLP and image recognition.
• Natural Language Processing: Focuses on computer-human language interaction, used in machine translation and speech recognition.
• Robotics: Involves the design and operation of robots for various industries, from manufacturing to healthcare.
• Expert Systems: Computer programs mimicking human experts’ reasoning and decision-making abilities, applied in fields like medical diagnosis and customer service.

Artificial Intelligence and the Arts

• AI art refers to visual artwork created using artificial intelligence (AI) programs.
• In the early 21st century, AI art tools became more accessible to the general public, extending beyond academia and professional artists.
• AI art has raised philosophical concerns related to copyright, deception, and its impact on traditional artists, including their incomes.
• In 2014, Ian Goodfellow and colleagues at Université de Montréal developed Generative Adversarial Networks (GANs), a form of deep neural network for creating new images.
• GANs operate with a “generator” to produce images and a “discriminator” to assess their quality.
• In 2015, Google released DeepDream, a program utilizing convolutional neural networks to enhance patterns in images, resulting in dream-like and psychedelic visuals.
• In 2018, Christie’s Auction House in New York hosted a significant auction of AI art, featuring “Edmond de Belamy” by the art collective “Obvious,” which sold for $432,500, far exceeding its estimate.
• During the 2020s, text-to-image models made substantial progress, reaching a level of quality comparable to real photographs and human-created art.
• In 2021, OpenAI’s DALL-E, powered by Transformer models from GPT-2 and GPT-3, demonstrated the ability to generate high-quality images based on natural language prompts.
• In 2021, EleutherAI introduced the open-source VQGAN+CLIP, which is based on OpenAI’s CLIP model.
• In 2022, notable developments occurred with the release of “Midjourney” and “Stable Diffusion,” contributing to a substantial growth in the field of AI-generated visual art.
• Also in 2022, artist RefikAnadol created an AI art installation at the Museum of Modern Art in New York, using the museum’s collection as a basis for the project.

Impact of Artificial Intelligence in the World of Arts

• AI accelerates the creative process and challenges traditional notions of authorship and authenticity.
• AI tools speed up art creation, with GANs generating realistic images in seconds.
• Image classification systems allow the rapid generation of multiple images.
• AI promotes diversity in art creation by overcoming linguistic, cultural, and geographic barriers.
• Increased diversity leads to greater representation and inclusion in the artistic community, amplifying diverse voices and perspectives.

Major Concerns

• Questions arise about the artistic value and authenticity of AI-generated copies, affecting the art market.
• Fear of standardization and homogenization of art due to AI’s algorithmic nature, potentially leading to mass-produced, similar artworks.
• Potential reduction in originality, diversity, and artistic value, making art less interesting and less valuable to the public.
• Standardization and homogenization can hinder creativity and innovation, making the art industry less appealing to new talent.
• AI’s efficiency in art creation raises concerns about a decrease in demand for human artists and fewer job opportunities.
• Replacing human artists can diminish the uniqueness and subjectivity of art, affecting individual expression and personal experiences.
• AI algorithms can carry biases and stereotypes, resulting in discrimination and injustice in art creation.
• The use of AI for unethical purposes, like creating fake news, raises ethical concerns.
• AI-generated images may challenge authorship and authenticity, affecting the value and meaning of these works.
• AI’s advanced nature can create technology disparities and limit access, potentially leading to inequality.
• Difficulty in accessing AI technology may reduce diversity in art production and exclude underrepresented groups from the industry.

AI and the issues related to Copyright Laws

• AI, including tools like ChatGPT and Midjourney, has prompted challenging questions regarding intellectual property (IP) rights.
• Generative AI enables individuals with limited creative skills to produce impressive content through text prompts.
• Concerns arise about potential copyright infringement when AI systems use copyrighted materials like texts and images for training data.
• Questions also revolve around copyright ownership for content autonomously generated by AI or with human inputs.
• In a recent case, Stephen Thaler vs. ShiraPerlmutter in the United States District Court for the District of Columbia, questions about copyright for AI-generated works have emerged.Mr. Thaler claimed ownership of the AI system ‘Creativity Machine,’ which autonomously generated a visual artwork.He applied for copyright protection, listing ‘Creativity Machine’ as the artwork’s author and intended to transfer the copyright to himself as the AI system’s owner.The U.S. Copyright Office rejected the application, emphasizing the absence of human authorship in the submitted work.
  • This case underscores the central question of whether AI-generated works can qualify for copyright protection.
• The main legal issue was whether an AI-generated work could be eligible for copyright.
• The court examined statutory provisions, past legal cases, and theoretical arguments for copyright.
• The conclusion was that copyright protection requires human creativity.
• The term ‘author’ as used in the Constitution and the Copyright Act specifically excludes nonhuman entities.

The Indian Context

• Mandatory AI use disclosure requirements and broader consultations haven’t been established by the Copyright Office.
• The 161st Report of the Department-Related Parliamentary Standing Committee on Commerce, issued in July 2021, is titled ‘Review of the Intellectual Property Rights Regime in India.’
• The report recommended a review of the Copyright Act 1957 and the Patent Act 1970 to encompass emerging AI technologies.

Global Perspectives on AI Art and Copyright

• In the United States, AI-generated artwork lacking “human authorship” is not protected by copyright.
• In the European Union, copyright law requires a work to be the “author’s own intellectual creation,” suggesting human involvement is necessary.
• However, certain EU judgments have recognized that creative choices in AI-assisted works can be significant, potentially leading to copyright protection.
• The European Commission has proposed a “four-step test” to qualify AI productions as a “work,” but legislative developments are ongoing.
• In Malta, copyright protection is granted to works created by natural persons, meaning human authors, and similar principles to the EU apply to AI-assisted works.
• The United Kingdom allows copyright for works generated entirely by AI, with reduced protection duration compared to human-created works.
• Ireland, India, and New Zealand also offer copyright protection to AI-generated artwork, with specific laws in place to address this emerging area.

Click to read more articles related to Artificial Intelligence

The Future

• The increasing use of AI in art blurs the line between human and machine-created artwork.
• Advances in computing power may make it challenging to distinguish human and AI-generated content.
• Future discussions should focus on protecting works created by intelligent algorithms with little human intervention.
• Copyright laws may need to establish exceptions for AI-generated content to ensure continued technological advancement.
• Granting copyright to those enabling AI operations is a sensible approach, promoting investment in technology.
• The potential debate about granting computers the status and rights of people is a separate issue.

What is the Way Forward?

• Exercise caution when extending traditional IP protections to AI-generated works, given potential negative consequences of monopolies.
• Acknowledge that conventional economic incentives for creators through copyright and patents may not directly apply to AI-generated content.
• Preserve the human-centric nature of copyright law and consider the role of human involvement in creative processes.
• Conduct a thorough assessment of AI-generated art and copyright laws, addressing specific challenges and implications.
• Encourage public dialogue and consultation on AI art and copyright for a balanced approach.
• Support further research on AI’s impact on the art world to inform legal and policy decisions.
• Collaborate internationally to establish a consistent framework for AI art and copyright.
• Be open to adapting copyright laws to accommodate AI art while safeguarding creators’ and the public’s interests.