Despite efficiency and cost reduction gains provided by the adoption of (IA), most legal organizations do not yet have a clear strategy for this technology. This is the main conclusion of Thomson Reuters’ Future of Professionas 2025, which shows a clear division between the offices and departments that advanced with structured planning and those who follow the adopting on a punctual or disorderly manner.
According to the survey, made with 2,275 professionals from the legal sector and areas such as compliance, audit and tax in several countries, only 22% of organizations have a visible AI strategy. Still, these institutions are twice likely to see AI -related revenue growth and 3.5 times more likely to extract critical benefits from technology.
In Brazil, the use of AI advances on different fronts, but also with challenges. “Legal work is being shaped by AI. Offices that do not adapt to the risk of being behind,” warns Steve Hasker, CEO of Thomson Reuters. According to him, there is an estimated amount of amount to estimated $ 32 billion in the US alone with the use of AI in the legal and accounting areas.
Technology
50 prompts to organize your financial life
Better time use
The study also points out that IMA professionals should save an average of five hours a week by 2025. This is the equivalent of an annual gain of up to $ 19,000 per person. In practice, it means more time to deal with complex tasks of higher value added to customers.
Exactly that is happening in the Office Marcus Valverde Lawyers Society. According to partner Marcus Valverde, the use of AI has been increasingly essential in the daily life of the bank. “Today, LegalTechs offer extremely personalized solutions for each type of work, allowing us to better channel our time to more strategic issues, leaving repetitive tasks with technology,” he explains.
Support tool
Even with the enthusiasm for automation, there is consensus among professionals in the area about the need for caution. “AI is a powerful tool, but it does not replace the lawyer,” says Antonio Carlos de Oliveira Freitas, director of the São Paulo Lawyers Association (AASP).
Continues after advertising
The entity has been investing in training and closely follows the regulatory debate. “It is necessary to be careful in the use of these tools, because we have seen situations where AI invented jurisprudence or brought wrong interpretations of the law. Therefore, use must be ethical and always supervised by the human,” says Freitas.
This concern is also shared by Matheus França, from Gaia Silva Gaede, who highlights the risk of biases and misuse of sensitive data inserted in the artificial intelligence system. “Despite operational gains, it is essential to implement AI with governance and responsibility,” he says.
This is exactly why the adoption of AI by the Attorney’s Office of the State of Sao Paulo is still incipient. According to Júlio Rogério de Almeida Souza, representative of the Attorney’s Office in the Data Management Committee, adoption is seen as inevitable. “But you can’t play confidential process information on unmistvalized systems. Responsibility for decisions remains from the prosecutor. However, systems need to ensure traceability and transparency,” he says.
Continues after advertising
Inequal Adoption in Justice
Within the judiciary itself, adoption is still unequal. “AI is already used in screenings, case law analysis and report production,” explains Daniel Marques, president of the Brazilian Lawtechs and Legaltechs Association (AB2L). He cites projects such as Logos, of the Superior Court of Justice (STJ), and supports the National Council of Justice (CNJ), which already act in dozens of Brazilian courts.
Still, the final decision remains with the professional. “Judges and lawyers should remain legal guardians. AI is support, not value judgment”, reinforces Marques.
No -return path
Bill 2338/23 is expected to regulate the use of AI in Brazil, including the governance part, advances in Congress. The text has already been approved by the Senate and, to become law, must also go through the sieve of the House of Representatives. The proposal classifies AI systems regarding risk levels for human life and threat to fundamental rights. It also divides applications into two categories such as Artificial Intelligence and Generative Artificial Intelligence.
Continues after advertising
By text, AI is defined as the machine -based system capable of, from a data set or information received, generate results such as content forecasting, recommendation or decision that may influence the virtual, physical or real environment. The Generative is defined as a model specifically intended to significantly generate or modify text, images, audios, videos or software code.
“It is essential that the use of AI is transparent and for the citizen to know when faced with machine -generated content,” says Freitas.
In offices and courts, the advancement of technology adoption is seen by experts as a great challenge that will require the consolidation of good practices and ensure that the advance is always aligned with the ethics, safety and quality of legal services.
Continues after advertising
Despite the enthusiasm for efficiency gains, Thomson Reuters warns that the greatest risk of organizations is not planning. Without a clear strategy, the use of AI can lead to more problems than solutions. “Companies need to align technology with their priorities. Only then will they get sustained innovation and relevance to their professionals,” says the company’s Global CEO.
