Introduction
AI has quickly become a crucial element of every corporation these days. Businesses are turning to AI systems to handle financial analysis, recruiting, market forecasting, and business decisions. AI has been commercialised, making technology a valuable asset for the corporate world and thus making it a popular choice in the industry. But, as AI becomes increasingly common, there are also significant legal issues that have come to light. Much of corporate/commercial law has developed in the context of human decision-making and identifiable responsibility structures. AI systems defy these assumptions by working in often opaque and complex ways, and autonomously. That has made it more difficult to hold accountabilities for actions that lead to financial losses, substandard decisions, or discriminatory output in cases involving AI systems.
This article presents that the commercialisation of AI has led to an emerging accountability gap in corporate governance and commercial liability systems. It explores governance issues around AI decision-making in the corporate landscape, contractual and liability issues relating to commercial agreements and the regulatory landscape of assigning legal liability in AI-driven commerce.
Commercialisation of AI and the Changing Corporate Landscape
The commercial value of AI has risen significantly in the last decade. AI systems are becoming a key differentiator in the business landscape, helping large technology firms, financial institutions, e-commerce platforms, and multinational companies operate more efficiently and effectively. AI has become a part of commercial operations like automated customer interactions, predictive analytics, fraud detection, and forecasting market trends and supply chain management. The commercialization of AI has also led to a fierce competition between big companies to establish themselves as the leaders in the AI space. AI models, datasets and algorithms are becoming companies’ valuable intellectual and commercial property.
The commercial use of AI in India mirrors the nation’s digital transformation agenda, a trend seen in the rising adoption of AI for business applications. AI has been recognised as a key driver of economic development through government initiatives like the NITI Aayog’s National Strategy for Artificial Intelligence. AI-powered systems are gaining significant traction in businesses, especially in the finance, healthcare and technology industries, for operational and strategic needs. But as AI has rapidly become part of the business mainstream, detailed legal regulation has lagged. Unlike commercial products, AI systems can often use machine learning algorithms that can adapt and learn over time. This gives rise to uncertainty about predictability, transparency and responsibility. The lack of clear rules in the area of governance and accountability for AI presents a high risk for corporations, consumers and commercial actors. The legal challenges related to commercialisation of AI are especially reflected in the corporate governance and commercial liability domains. The pressure is mounting as businesses increasingly turn to AI systems for commercial tasks, forcing traditional legal concepts to evolve.
Corporate Governance Challenges in AI-Driven Decision-Making
Corporate Governance is one of the biggest legal implications of AI commercialisation. Corporate governance is a system that guides and manages companies. Directors are required to exercise fidelity, care and duty to act in the best interests of the company. As AI systems become a more integral part of corporate decisions, questions are emerging about what this means for the role of directors and whether they can rely on these systems and still have accountability.
As per the Companies Act 2013, a company’s directors are required to demonstrate reasonable care, skill and diligence with respect to management of company affairs. Under section 166 of the Companies Act, directors have a duty to “act in good faith” and to “exercise independent judgment.” As AI systems become more prevalent, there is a tension with these responsibilities, as directors might be more relying on automated systems to make strategic and financial decisions. AI tools are being used by corporations more and more to analyse consumer behaviour, evaluate investment risks and assess market trends, for example. While AI can streamline processes, using automation too heavily can reduce the human element in decision-making. The harmful commercial outcomes may be an attempt to justify it by claiming that the commercial outcome was arrived at through an AI system generating the relevant commercial outcome. This raises a major accountability issue because typically fiduciary duties involve one or more humans making an active decision. It becomes even harder to pinpoint the responsibility if the AI systems are misaligned and result in discriminatory or harmful outcomes. For example, some AI recruitment systems have been criticized for disadvantaging female applicants because they were trained on historically male-dominated hiring data. When a corporation is based on a biased AI system that generates discriminatory results, there is an issue of the extent to which the directors exercised oversight. It is crucial to have good corporate governance, corporate responsibility and attention to fiduciary duties, as brought out in the Supreme Court’s findings in Tata Consultancy Services Ltd. v. Cyrus Investments Pvt. Ltd. The principles of responsible governance are applicable to the corporate use of AI, though the case was not directly regarding AI. In a similar vein, the Supreme Court highlighted the need for transparency and accountability in corporate actions in N. Narayanan v. Adjudicating Officer, Securities and Exchange Board of India. In a commercial environment powered by AI, these ideals matter more because algorithms can be opaque, leading to a reduction in accountability.
The other governance problem is of explainability. There are also many tools that use very complicated machine-learning mechanisms, whose calculations are not easily understood, even by the programmers. This “black box” nature of AI poses complications to directors trying to prove they’re fulfilling fiduciary responsibilities. When directors are not able to get to the bottom of an AI system’s decision-making process, determining the reasonable oversight is challenging. AI commercialisation can thus lead to governance regimes in which companies increasingly rely on a lack transparently operated systems for decision making by the directors. Rules and principles of corporate governance might then need to be reinterpreted to make sure that directors are not shielded from taking any accountability by relying too heavily on the algorithmic systems.
Commercial Liability and Contractual Challenges
The commercialisation of AI has also created substantial challenges within commercial law, particularly regarding contracts, liability allocation and defective AI outputs. AI systems are commonly supplied through licensing agreements, software-as-a-service arrangements and commercial partnerships. However, traditional commercial law frameworks often struggle to address the unique characteristics of AI technologies.
One major issue concerns contractual liability for inaccurate or harmful AI outputs. AI systems frequently operate through predictive models that may generate probabilistic rather than guaranteed outcomes. Commercial contracts traditionally assume a relatively predictable standard of performance. AI systems, however, may produce inaccurate or evolving results depending upon data quality, training methods and machine-learning processes. For example, a corporation may purchase an AI-powered financial analysis system that generates inaccurate predictions resulting in significant commercial losses. Determining liability in such situations becomes difficult. The software developer may argue that the AI system was merely predictive and not intended to guarantee accuracy. The purchasing corporation may nevertheless argue breach of warranty, negligence or misrepresentation. The Indian Contract Act 1872 continues to govern many contractual relationships involving AI technologies. Sections relating to misrepresentation, consent and contractual obligations remain relevant to AI licensing agreements. However, the Act was not designed to regulate autonomous algorithmic systems capable of evolving through machine learning.
Another important issue concerns limitation of liability clauses within AI contracts. Technology providers frequently include contractual provisions limiting liability for defective AI outputs. Such clauses attempt to shift commercial risk onto users or purchasers. However, as AI systems increasingly influence critical commercial decisions, excessive contractual limitations may create unfair allocation of risk. Product liability also presents significant legal uncertainty. The Consumer Protection Act 2019 introduced product liability provisions within Indian law. However, applying product liability principles to AI systems remains challenging because AI outputs are dynamic rather than static. Traditional product liability frameworks assume identifiable defects within fixed products. AI systems, by contrast, continuously evolve through learning processes.
The issue of data ownership further complicates commercial AI relationships. AI systems rely heavily upon large datasets for training and operational effectiveness. Businesses increasingly compete over access to commercially valuable consumer and operational data. Questions regarding ownership, licensing and commercial exploitation of such data remain legally contested. Privacy regulation also intersects with commercial AI liability. The Digital Personal Data Protection Act 2023 has obligations for the processing and use of personal data in India. Consumer information shared with AI systems can introduce risk of illegal access to consumer information, consumer profiling, and/or illegal commercial use. Non-compliance with data protection requirements can lead to legal sanctions and damage reputations for companies.
Globally, AI systems are increasingly recognized as being commercial risks. The EU AI Act is one of the first comprehensive efforts to classify and regulate AI into risk categories. The Act will establish new obligations on high-risk AI systems, especially in terms of transparency, accountability, and human oversight. The law is only in effect in the European Union, but it will certainly have effects outside the EU as to many multinational companies, operations are spread across national borders. The commercialization of AI has thus revealed certain weaknesses in existing commercial law. Current contractual and liability regimes are not well suited to the nature of autonomy, opacity and constant change that typically defines systems.
The Emerging Accountability Gap and Regulatory Challenges
The most significant legal consequence of AI commercialisation is the emergence of an accountability gap. AI systems diffuse responsibility across multiple actors including developers, vendors, deployers, directors and end-users. This fragmentation makes it increasingly difficult to determine legal responsibility when commercial harm occurs.
Traditional corporate and commercial law frameworks generally assume identifiable human actors exercising direct control over decisions and transactions. However, AI systems contradict this assumption as their outputs can be the result of many contributors in complex machine-learning processes. For example, if an AI tool, a company has developed and implemented is using data that results in biased hiring practices, the software developer, the company using the tool as well as the creators of the data or the those in charge of implementation could be responsible. Current laws and regulations frequently do not have clear procedures for determining who should be held accountable for these interrelated actors. This accountability blind spot is a major worry, especially as AI systems grow more prevalent in high-risk industries like banking, healthcare and financial services. In absence of comprehensive regulatory standards, the corporations may become over dependent on the automated systems and shareholders may evade the responsibilities. In India, there is no dedicated standalone AI law at this moment, similar to the EU AI Act. The current governance is mostly sector-based, while the existing data protection law and general corporate & commercial law principles are mostly relied upon. This flexible approach could foster innovation, but it can also lead to regulatory uncertainty. As the importance of AI governance grows, scholars have increasingly called for a multi-layered approach of corporate accountability, contractual regulation and public oversight. While existing legal frameworks might not apply directly, they may need to be expanded to accommodate the unique elements of AI systems, such as their potential to possess agency and responsibility beyond human expectations.Existing legal frameworks may not fully capture the nature of AI systems, and they might require modifications to account for their unique features and capabilities, including agency and responsibility that extend beyond human expectations.
This commercialisation of AI, in turn, calls for a rethinking of the concept of accountability in technologically mediated markets. In absence of more effective governance regulations and liability mechanisms, companies can keep going business as usual in a messy regulatory system which is increasingly hard to define the responsibilities.
Conclusion
With the continued growth and penetration of AI into the commercial sector, robust governance principles and frameworks, unambiguous liability provisions, and a more robust regulatory system will become increasingly important. If legislation does not come into effect, AI commerce could remain in a patchwork and ambiguous regulatory landscape.
THIS ARTICLE IS WRITTEN BY ZOBIA FATIMA FROM QUEEN MARY UNIVERSITY OF LONDON
REFERENCE :
Companies Act, 2013, s 166
Indian Contract Act, 1872
Consumer Protection Act, 2019
Digital Personal Data Protection Act, 2023
NITI Aayog, National Strategy for Artificial Intelligence #AIForAll (2018)
Tata Consultancy Services Ltd. v. Cyrus Investments Pvt. Ltd. (2021) 9 SCC 449
N. Narayanan v. Adjudicating Officer, Securities and Exchange Board of India (2013) 12 SCC 152
European Union, Regulation Laying Down Harmonised Rules on Artificial Intelligence (Artificial Intelligence Act), European Parliament Legislative Resolution, 2024
Surden, Harry, “Artificial Intelligence and Law: An Overview” (2019) 35 Georgia State University Law Review 1305
Pasquale, Frank, The Black Box Society: The Secret Algorithms That Control Money and Information (Harvard University Press 2015)
Cary Coglianese and Alicia Lai, “Algorithmic Governance and Accountability” (2021) 95 Public Administration Review 761
