As a Founder and Managing Director, Bikram plays a key role in the day to day operations and formulating strategies for smooth execution of research assignments. Bikram is also responsible for client acquisition and client engagements in Asia pacific region. Bikram helps global technology companies monetize their IP assets through automated tools and helps them generate more value out of their IP. He has the expertise of assisting clients on a wide range of technologies such as Semiconductors (circuits and processes), Wireless Technologies, Memory Devices, Cloud Computing, and Consumer Electronics.
Bikram has been advising several Global 500 clients in effective patent portfolio optimization and competitive benchmarking analysis to help them understand their position in the industry. Bikram has extensive experience in assisting clients on high stake US, Japanese and Chinese patent litigation cases. Recently, Bikram has been actively involved in advising clients on new product development strategies and technology Landscape studies.
The swift development of Artificial Intelligence (AI) has increased adaptability across industries like healthcare, life sciences, financial services, and telecommunication. Although unregulated, AI is significantly affecting the Intellectual Property landscape. Against several decades of enthusiasm and waves of setbacks, AI’s emergence has unfolded great application uses that are solving complex business problems today. With its rapid growth, it is anticipated that conflicts between IP Protection and enforcement issues will increase.
Within the intersection of AI and IP lies legal regulations and technological innovations. To understand this interplay between law and artificial intelligence, one needs to gain an in-depth overview of the AI and IP shared space that includes AI-generated artistic creations and their ownership, patent protection for AI inventions, liability for AI-generated harm, data protection and privacy, use, and benefit from open-source AI software, licensing AI technology and infringement of AI-generated works, under the legal and regulatory framework.
Ownership of AI-generated works – Patent protection for AI inventions
Who owns content generated by AI? Who owns the rights to the output of AI systems, such as music, or images? The ability to create incredibly powerful images conditioned on detailed text descriptions, are taking over the internet audience by the storm. When the artwork, however is generated, can you say you own it? Or say that AI owns it? In that case, is “AI” an artist, as stated by the law? Although these artistic creations are perceived to be human-like owing to the detailed prompts, they are anything but human. These are language models, solely trained for receiving prompts and generating images.
The question remains – who is the owner of AI-generated works? How are the Intellectual Property rights designed around AI-prompt marketplaces? Well, patent and IP issues in AI modelling is slightly foggy. US Intellectual Property regime has not openly recognized any form of non-human creation or invention. Parts of Asia considers AI-generated literary, musical or dramatic works as the creation of the author who puts the prompts and ordering together. Hence AI- generated works in countries like Hong Kong are protected by copyrights. But the statutory regime does not contain any section covering specifically AI.
Rapidly emerging AI innovations has significantly gained momentum in India and ranked 8th in terms of AI patent filing over the last decade. Interestingly, artworks and inventions generated by AI are observed similar to CRI (computer-related inventions) in India. However, there have been concerns over interaction between IP and web-scraping for AI use. As technology has advanced over the years, statutory regulations have not been able to keep pace with the changing landscape of AI ecosystem.
Unfortunately, patent protection and enforcement law for AI-generated inventions is still ambiguous. Statutory laws and TRIPS agreement requires the creator/inventor to be a human being or a legal individual. AI conforms to none. This raises serious concerns on the patentability of AI creations. Currently, there are no directions from the courts that clearly states ownership rights in AI creations.
Data protection and privacy – Intersection with AI
Winning AI strategies ensure that businesses are cognizant of mitigating risks arising specifically from data privacy issues. It is important to understand the intersectional context between AI and privacy legislation. While the possibilities in the AI world are tremendous, concerns about misuse of data and its privacy are always at risk. Collection and analysis of data becomes one of the first touchpoints where AI and data privacy intersect. This information is usually sourced from social platforms and websites that may contain sensitive financial and medical data, person’s political views and other personal details. Businesses are wary about technology models obtaining and processing their data, which can also be sold off to third parties.
Another way that they intersect is through machine learning algorithms. There are risks of perpetuating biased interpretation and data discrimination as AI systems train and adapt to data set, which may not be the correct representative outcome. These are predictive behavioral analysis or contextual analysis of a race or an ethnic group.
Given the complexity of data privacy issues in AI, businesses have to implement strong data privacy policy and practices that are statutorily compliant. Some of the key measures adopted by organizations include:
- Define AI code of ethics – Develop a strong set of transparent principles and techniques in the fabric of your product that reflects the commitment and responsible use of AI technology. Let the backend teams review every AI/ML product with the ethical lens as defined by the business.
- Create a group of domain experts and mentors – Form a group of AI/ML experts to guide through decisions on design and development. Diverse perspectives and cross-industry experience-sharing from legal experts, public policy, ethics and compliance task force will help review potential issues across product lifecycle stages.
- Train your team to build ethical products – Identifying existing infrastructure to operationalize data and AI ethics, responsibly mitigate risks and create an AI ethical product framework and release it soliciting honest reviews.
European Union has implemented The General Data Protection Regulation (GDPR), one of the toughest privacy laws in the world. It applies to any organization that collects, processes, and stores personal data of EU citizens, irrespective of the organizational location. GDPR aims at giving individuals enhanced control over their personal data, by being more transparent and allowing them right to request access, rectification, erasure, and portability of their personal data, adding a layer of complexity to AI-driven businesses. Such stringent regulations protect data privacy and establish security-conscious workflows.
India currently lacks an express law on data protection, but the IT Act of 2000 and IT Rules of 2011 provide some guidelines on the handling of sensitive personal data or information. However, these rules have limitations, as they only apply to corporate entities in piecemeal and do not cover government agencies.
Open-source AI software: How businesses use and benefit from OSS by respecting IP rights?
Permissible open-source software are legal tools for mass market adoption. Sounds great? Most certainly yes, it does. How fascinating it is to be able to cooperate and share information as AI developers adopt open-source platform to build on others’ work, share their own contributions and rapidly scale up advancements. In that case, open-source software usage must take off and disrupt the world of innovation. But reality, however, is far from ideal. Because using open-source has certain catches and vulnerabilities and exposing businesses to data breaches is one of them.
Read: Pros-Cons of using ChatGPT for IP & legal firms
Open-source license terms may mandate the distribution of source code for modified versions of open-source programs, which could also extend to proprietary software that incorporates this open-source software. Vulnerabilities are not unique to open-source software and even the most confident coder cannot claim their code is perfect.
Companies can use and benefit from open-source AI software while respecting IP rights by ensuring that they comply with the terms of the open-source license and by avoiding any infringement of third-party IP rights. They can also contribute to the open-source community by making their own improvements and modifications to the software available under the same open source license. Open-source leverages IP as a tool for diffusion of innovation, allowing companies to extract ROI by focusing on distribution and attraction, and tapping into adjacent connected markets, such as cloud computing, rather than direct licensing of IP. Traditional IP strategies enforce their right to exclude others to prevent free-riding or imitation. Organizations therefore benefit from relying on these IP strategies.
From a market competition standpoint, open-source dynamics play a complex role. While it promotes broader access and participation on one end, it may also lead to firms competing for network effects and market tipping. Leading AI companies employ a hybrid strategy of accumulating patents and investing in the open-source community. The debate on AI-related patents is divided, with some advocating for patent protection to encourage innovation, while others believe that it discourages innovation by privatizing basic AI elements and leading to costly litigation.
Licensing AI technology: How can companies license and commercialize their AI technologies while protecting their IP rights
While AI uses and benefits are growing manifold to lend a competitive edge to businesses, its challenges also increasing exponentially. Key issues arise in licensing AI technological innovations while protecting their IP ownership rights, infringement, data use, privacy and legal compliances.
One of the fastest ways to license AI capability is through a third-party provider. AI licensing may come in the form of an on-premise license or a Software as a Service (SaaS) solution in the cloud. However, the output, value, and performance of an AI system cannot be accurately predicted, making it challenging to make strong promises. Moreover, machine learning capabilities improve over time so it is not prudent for businesses to test the product performance at the onset and work over licenses.
How do businesses license their technologies under uncertainties as these? Many businesses are turning to a collection of AI providers to test the waters through a low-risk way of proof-of-concept arrangement. It is a short-term agreement that allows a company to test and a supplier to prove the value of an AI product or service.
The next step for a business is to obtain AI license from a provider. At this stage, businesses should ensure that they meet the standard requirements for license and SaaS agreements while focusing on these four unique areas.
- Legal Compliance – Regulators’ increasing focus on AI have not been able to fully protect their patents as on day. Most advanced laws have not caught up to AI yet. To ensure regulatory compliance, a license agreement for an AI may need to include provisions that allow for regulatory examination that meets all regulatory requirements and prevent potential legal issues.
- IP Protection Rights – Current IP and patent laws in technologically advanced countries as US, are designed to protect human creativity, and may not be fully applicable to AI. Contractual protections are therefore crucial for capturing and preserving the value of AI investments, and these protections should be established before beginning an AI project.
- Data Use and Privacy – When using data to fuel AI systems, businesses must comply with any contractual requirements set by third-party data suppliers. To avoid potential legal issues related to data usage, businesses should insist on learning about the level of legal and regulatory diligence that has been conducted on data usage in AI systems.
- Embedded AI licenses – Embedded AI systems are becoming more common in consumer products, with voice-controlled devices, autonomous vehicles, and other personalized recommendations. Additionally, businesses are incorporating AI into their operations and business-to-business offerings. They work with AI providers closely to stay ahead of competition.
Conclusion
Patents are effective in protecting AI-related intellectual property, especially since independent creation does not serve as a defense against patent infringement. AI is no longer just a futuristic concept, but has real-world applications that require proper protection from infringement and misappropriation. Having a comprehensive AI protection strategy in place can help organizations enforce their intellectual property (IP) rights effectively and prevent unlawful conduct. Therefore, organizations should develop a robust AI protection strategy to safeguard their intellectual property rights, keeping in view the legal, privacy and other IP protection aspects discussed above in detail.
