On 11 September 2024, I shared insights on the impact of artificial intelligence (AI) on patent drafting at EuroCham Singapore's webinar, held in collaboration with the Southeast Asia IP SME Helpdesk. Experts from different sectors gathered to discuss how AI is reshaping the intellectual property (IP) landscape, focusing on copyright, patent drafting, and trade finance. Below, I provide a summary of the main points discussed during the event.
AI and Copyright: Legal Implications and Ethics
Max Ng from Gateway Law Corporation presented on the legal challenges AI-generated content poses for copyright. AI tools like ChatGPT and Midjourney can generate large volumes of text, images, and music, closely mirroring human-created works. However, courts have ruled that copyright applies only when a human creates the work.
A significant case discussed was Steven Thaler's attempt to list an AI system, DABUS, as the inventor in a patent application. Courts in the United States and other jurisdictions rejected the application, affirming that only natural persons can be inventors. This decision reinforces the principle that while AI can assist human creativity, it does not meet the legal standard for genuine authorship.
I believe that understanding this distinction is crucial. AI can enhance human productivity and creativity, but it does not and will never make human involvement in authorship redundant. There are these assumptions that are shared among religious members of the AI community: life is digital as well as personal and digital processes, if given enough computing power, and if given sufficiently complex programming, which means programming by other computer programs, can and will become the same as personal knowledge. For believers in God, there is the issue that this makes the concept of man being created in the image of God incorrect. Will a computer exercise its judgment based on some form of ethics? The term "ethics" has to do with responsibility. But responsibility to whom? If individuals are responsible to God, and if there is a final judgment, does this also apply to computers? Will they go to hell?
There are two more assumptions that are shared among religious members in AI: knowledge is power. So, greater knowledge is greater power. And in the end, the robots will inherit the earth. And digital processes, over time, will gain knowledge and then even wisdom.
This is the theology behind that AI religion: man is the present supreme god, and he then creates digital processes that in turn create life. This newly created life possesses enormous knowledge and therefore possesses enormous power. Followers of the AI religion may have started to jump when they read the word "theology" for what they are promoting. It may not be called theology, but it is one.
Copyright Infringement and the Fair Use Defense
Max Ng demonstrated that the legal complexities of copyright infringement in AI-generated content are significant, with cases like Getty Images' lawsuit against Stability AI highlighting these challenges. Getty alleges that Stability AI used its copyrighted images without licenses to train its models, thus allegedly breaching copyright laws.
The legal landscape remains unsettled, with "fair use" potentially applicable in some cases. However, its success depends on the nature of the content and its intended purpose. Businesses that use AI-generated content must be aware of these legal risks and establish mechanisms to protect their IP rights.
AI and Patent Drafting: A Demonstrative Approach
During the webinar, I demonstrated a custom AI workspace used at Schweiger and Partners that automates various aspects of patent drafting, including the generation of claims and figure descriptions. Through a live demonstration, I showed how the AI setup could produce a draft patent application in minutes, a task that typically takes several hours. This efficiency highlights AI's capacity to double productivity while upholding high standards of accuracy and quality.
Key Security Risks in AI-Enhanced Patent Drafting
Despite its advantages, AI use in patent drafting comes with significant security risks:
- Data Ownership and Security: Some AI platforms may store or use input data for training, which could unintentionally expose confidential information. In patent drafting, this is particularly concerning, as it may compromise the integrity of a patent.
- Data Breaches: AI systems integrated into IT infrastructure are vulnerable to cyber-attacks. A breach could result in the exposure of sensitive client data, causing reputational damage to the firm. It is crucial to recognize these risks and implement adequate security measures to mitigate them.
- Intellectual Property (IP) Theft: Poor data management practices can lead to IP theft, as unauthorized access to sensitive data could compromise client rights or disclose proprietary inventions.
Strategies for Protecting Client Information
To mitigate these risks, I recommend several strategies for managing AI tools securely in the patent drafting field:
- Select Secure AI Platforms: Opt for AI tools that meet stringent security standards. Avoid free or open-source options, as they may not guarantee data privacy. Instead, choose enterprise-level platforms with transparent data protection policies and clear ownership rights.
- Use On-Premise AI Solutions for Sensitive Data: For highly sensitive patents, such as those in the defense area or in pharmaceuticals, on-premise AI tools hosted on internal servers are advisable. This approach ensures full control over data and minimizes risks associated with cloud services.
- Implement Legal and Technical Safeguards: At Schweiger and Partners, we use non-disclosure agreements (NDAs) with AI vendors to legally protect client data. Additionally, we employ technical measures, such as encryption, two-factor authentication, and firewalls, to prevent unauthorized access.
- Regular Staff Training: Human error is a significant risk, even with advanced technology. Ongoing staff training is essential to ensure that everyone understands the secure use of AI tools and knows how to respond if a breach occurs.
AI in Trade Finance: Balancing Efficiency and Risk
Jose Lopez, Director of Digital Strategy at Flow Capital Management, shared insights into how his firm uses AI for predictive analysis in trade finance. By analyzing large datasets, including weather patterns, port schedules, and vessel traffic, their AI system optimizes cargo delivery schedules. This approach ensures that vessels arrive at the port precisely on time, maximizing efficiency and allowing the firm to offer more competitive services, even in a sector where margins are tight. This illustrates how AI can transform not only internal processes but also an entire business model, making it more adaptable and resilient.
However, Lopez emphasized the importance of risk management when integrating AI. The use of extensive datasets and AI models requires attention to data integrity and IP rights. At Flow Capital Management, they ensure that AI models are trained on legally obtained datasets and implement protocols to safeguard confidentiality and maintain accuracy.
Preparing for the Future: The Role of AI in the IP Sector
The future of AI in the IP sector presents both opportunities and challenges. AI has the potential to revolutionize industries by enhancing efficiency, speed, and precision. However, professionals must also understand its limitations and risks. As an IP professional, I emphasize the need to stay informed and adaptable.
AI will not replace professionals; rather, it will enhance the capabilities of those who know how to use it effectively. By selecting appropriate tools, implementing robust safeguards, and investing in training, businesses and patent professionals can effectively harness AI's capabilities while ensuring the security and integrity of their work.
For those interested in further exploring AI's potential in the patent drafting field, I offer tailored training sessions to help professionals integrate these tools efficiently and securely. My objective is to empower others to improve productivity and quality while upholding rigorous data protection standards.
IP Lawyer Tools by Martin Schweiger
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.
