Intellectual Property and Artificial Intelligence: Does the Content Your Company Creates on Platforms Such as Claude, ChatGPT, and Manus Really Belong to You?

The New Legal Dilemma for Companies in the Era of Generative AI

The massive adoption of generative artificial intelligence tools has radically transformed Brazilian corporate routine in recent years. Advertising texts, programming code, institutional images, financial spreadsheets, marketing strategies, and even technical opinions are now produced, partially or entirely, with the assistance of platforms such as Claude, ChatGPT, Manus, Gemini, and Midjourney. This movement, although representing significant productivity gains and reductions in operational costs, brings with it a complex legal issue still poorly understood by the national business community. Does the content generated by these tools really belong to the company that produced it?

The answer, contrary to what many imagine, is far from obvious. It involves the interpretation of the Brazilian Copyright Law, the terms of use of foreign platforms, the General Data Protection Law (LGPD), and international regulatory trends that have been influencing the domestic legal system. For managers, partners, and administrators, understanding these nuances has become essential to protect intangible assets, avoid litigation, and ensure legal certainty in everyday business operations.

The Brazilian Legal Framework of Intellectual Property in Relation to Artificial Intelligence

Law nº 9.610, of February 19, 1998, known as the Copyright Law, is the main legal instrument regulating the protection of intellectual works in Brazil. Its article 11 categorically establishes that the author is the natural person creator of a literary, artistic, or scientific work. This premise brings a fundamental legal consequence, since Brazilian legislation does not recognize, at least until the present moment, the authorship of works produced exclusively by machines or automated systems.

When a company uses Claude to draft a report, ChatGPT to develop an advertising text, or Manus to automate data analysis, a legal gray area arises. If the content was entirely generated by artificial intelligence, without significant human creative intervention, it may not be considered a work protected by copyright. This means, in practice, that such material could, in theory, be used by third parties without the company having legal grounds to prevent such use, a situation that represents a relevant equity risk.

On the other hand, when there is substantial human contribution in the process, whether in the elaboration of sophisticated commands, in the technical review of the generated material, in the careful curation of content, or in the creative combination of different results, it is possible to sustain copyright protection. The key criterion becomes the degree of intellectual involvement of the human contributor in the creative process, according to doctrinal understanding that has been consolidating in comparative law.

Platform Terms of Use and Their Contractual Implications

Another aspect frequently neglected by companies refers to the terms of use of artificial intelligence platforms. Each provider establishes its own rules about the ownership of the generated content, and these clauses usually have contractual validity even in Brazil, by force of the principle of autonomy of will and articles 421 and following of the Brazilian Civil Code.

OpenAI, developer of ChatGPT, provides in its terms that the user holds the rights over the generated outputs, with the caveat that identical outputs may be delivered to other users in similar queries. Anthropic, creator of Claude, adopts similar logic, with specific restrictions regarding the use of content for training competing models. Platforms such as Midjourney present distinct regimes between paying and free users, with the latter potentially having considerably limited rights over the generated images.

This contractual diversity requires redoubled attention from companies, especially those that produce commercially valuable content from these tools. Ignoring the terms of use may result not only in the loss of rights over the produced material, but also in exposure to litigation involving contractual breach, undue use of third-party intellectual property, and even violation of confidentiality clauses signed with clients or business partners.

Business Risks and the Protection of Confidential Information

Beyond the question of ownership, the corporate use of generative AI tools raises serious concerns related to confidentiality and to the General Data Protection Law, Law nº 13.709/2018. When employees insert sensitive information into prompts, such as client data, trade secrets, internal strategies, financial information, or protected personal data, these elements may be stored, processed, or eventually used for the training of models, depending on the platform and the contracted plan.

The Superior Court of Justice has reinforced, in several recent rulings, the objective liability of companies for inadequate processing of personal data, with cumulative application of administrative sanctions from the LGPD and indemnities for collective moral damages. In this context, the absence of clear internal policies on the use of artificial intelligence may configure business negligence, with serious equity and reputational consequences, especially in cases of security incidents involving strategic information.

Companies operating in regulated sectors, such as financial, legal, healthcare, and technology, face even higher risks. The inadvertent leakage of confidential information through prompts in public tools may characterize breach of professional secrecy, violation of contractual clauses with clients, and even infringement of sector regulations, such as the resolutions of the Central Bank, the Securities and Exchange Commission, and the professional councils responsible for the supervision of each category.

International Jurisprudence and Brazilian Regulatory Trends

International jurisprudence has offered important guidelines on the intellectual property of content generated by artificial intelligence. In the United States, the Thaler versus Perlmutter case, judged by the U.S. District Court in 2023 and confirmed on appeal by the U.S. Court of Appeals in 2025, established that works created exclusively by autonomous AI systems, without significant human intervention, cannot be registered with the U.S. Copyright Office. Decisions with similar reasoning have been issued in the United Kingdom, in the European Union, and in China.

In Brazil, Bill nº 2.338/2023, known as the Legal Framework for Artificial Intelligence, is in advanced processing in the National Congress and promises to bring greater clarity to the subject. The text under discussion provides specific rules on accountability, algorithmic transparency, user rights, and the treatment of data used in model training. Although still pending final approval, its provisions already reflect the doctrinal and jurisprudential understanding that has been consolidating, both nationally and internationally.

The Administrative Council of Tax Appeals, known as CARF, has also begun to address tax issues involving expenses with AI platforms, especially regarding the deductibility of expenses with services contracted from foreign companies and the eventual incidence of taxes such as IRRF, CIDE, and PIS/COFINS-Importation. These aspects, frequently ignored by managers, may generate relevant tax assessments for companies that intensively use such tools in their commercial operations.

Best Practices and Business Strategies for Risk Mitigation

In view of this complex scenario, attentive companies have adopted a series of best practices to mitigate legal risks related to the use of generative artificial intelligence. The first of them consists of the elaboration of clear internal policies, which regulate the use of these tools by employees, establishing what may or may not be inserted in prompts, which platforms are authorized, and which procedures must be observed to guarantee the confidentiality of business information.

The contracting of enterprise plans, instead of the use of free or personal versions, usually offers greater contractual protection, with specific clauses of confidentiality, non-use of inserted data for training, and even indemnification against third-party claims related to intellectual property. Companies that produce content of commercial value from AI must adequately document the creative process, evidencing substantial human contribution, in order to strengthen eventual claims for copyright protection in litigious contexts.

Contractual review with clients, suppliers, and business partners also becomes fundamental, especially in sectors where confidentiality is an essential element of the relationship. Clauses that regulate the use of AI in the execution of services, with transparency about which tools will be used and which safeguards will be adopted, confer legal certainty to all involved parties and significantly reduce the risk of future conflicts.

The Importance of Preventive Legal Counsel

The question about the ownership of content generated by artificial intelligence tools is far from being merely academic. It is a matter of deep business relevance, with direct implications on intangible equity, regulatory compliance, personal data protection, and civil liability. The answer to the initial question, therefore, depends on a multifactorial analysis that involves Brazilian legislation, the contractual terms of the platforms, the degree of human contribution, and the constantly evolving regulatory context.

Companies that wish to fully take advantage of the benefits of artificial intelligence, without exposing themselves to unnecessary legal risks, need to adopt a proactive posture. This involves a critical understanding of the applicable legal framework, the implementation of robust internal policies, the constant review of operational practices, and the monitoring of regulatory trends emerging on the horizon. Specialized and preventive legal counsel proves especially valuable in this context, contributing to the construction of a consistent strategy that combines technological innovation and legal certainty, two indispensable pillars for business competitiveness in the contemporary environment.