Artificial Intelligence is generally defined as an application of technology that is aimed at accomplishing tasks that are usually characterized by their requiring human intelligence. It entails the creation of applications that are endowed with capabilities to solve problems, think, learn, and even perceive similarly to human beings. These systems incorporate the use of formulas and data to assess the information and then come up with good decisions.

AI can be broadly categorized into two primary types: The first kind is Narrow or Weak AI, which is developed for some particular functions or fields while being useless outside those fields –, for instance, voice assistants like Siri or Alexa; in contrast, General or Strong AI is a more complex type that is theoretically capable of solving any 'intellectual' tasks that people can. Currently, there is still active discussion and development in the field of strong AI.

Thus, AI solutions are integrated into industries present in healthcare, finance, retail, manufacturing, transportation, media, and education. From its adoption in the legal fraternity, it has boosted the competencies of legal practitioners. AI-based tools cover activities like document review, due diligence, contract management, and even predictive analysis. They can excel at big data handling, filtering useful information from unimportant information, and saving much time and money.

They also undertake legal research and study case laws together with legal precedents as a way of helping legal professionals in the determination of cases and formulation of legal strategies for litigations. Also, such technologies help in contract management by extracting terms, clauses, and provisions of contracts, as well as risk assessment, compliance, and even custom contract generation. These technologies optimize the execution of procedures, most of which would take considerable time to accomplish.

Recent trends in the use of AI in the Indian legal framework include the Supreme Court implementing a state-of-the-art tool that adapts information receipt and passes it to the judges for determination even though the tool is not involved in determining any case. In Jaswinder Singh v. State of Punjab, the High Court of Punjab & Haryana dismissed a bail petition resulting from charges related to the fatal offenses of culpable homicide and grievous bodily injury. The judge of the case also used the help of a generative tool to get a wider view of whether or not to grant bail in cases of cruelty.

The use of this term does not indicate the developing opinion of the case and the trial court never took into consideration these comments. The reference to Woods was made solely to get a broader perspective on bail jurisprudence where cruelty is involved.

Indian Law firms and legal practitioners have realized the use of new technology in their day-to-day operations. For instance, Cyril Amarchand Mangaldas (CAM) recently agreed with Kira Systems, a Canadian solution supplying software based on machine learning to work more effectively, accurately, and faster in delivering legal services.

Kira Systems leverages artificial intelligence to analyze contracts and other legal paperwork to extract clauses and other data with the help of machine learning models for various deals for different types of law practice. To the same effect, other international law firms like Clifford Chance, Baker & Hostetler, Linklaters, and Pinsent Masons have adopted AI in their workflow procedures.

India's Judicial System has integrated modern technology into its working processes as well. The National Judicial Data Grid (NJDG) is a stunning instance of how technology supports the workings of the legal system of India. Legal prediction, case analysis, and automation procedures such as e-notices and e-summons may help release the pressure on the judicial system. Technology has also played a huge role in ensuring virtual courts via e-courts and VC tools as seen with the Live Transcription Project pursued by the Supreme Court of India.

The Indian government has expressed concern about the promotion of the use of advanced technology and its ethically sound application and launched the National Artificial Intelligence Strategy in 2018. Indian legal tech start-ups are present and active in the market creating innovations for the legal industry. These start-ups build applications that, using machine learning and NLP, discover and distill legal contracts, increasing the chances of finding the right information and the speed of researching it.

The state-of-art technology has various possibilities for improving society in different fields like health, learning, transport, and amusement. But like any other AI, it has its pros and cons, problems, and opportunities where some of the issues may be ethical dilemmas, violation of people's privacy, bias, discrimination, or even insecurity threats. To solve these problems, a Congress of international practitioners and data scientists presented a new, non-binding guide for creating these products.

The World Ethical Data Foundation (WEDF) with members of staff from Meta, Google, and Samsung spanning 25,000 members has a framework of 84 questions for developers at the initial development stage of a project.

Consequently, there is a need for specialized legislation for advanced technologies to regulate it, remove diverse inherent or acquired biases in decision-making processes, and solve the issues related to ethical dilemmas. Documents like white papers, guidelines, and policies address and regulate algorithmic impact assessments and delete algorithms' bias in jurisdictions like the UK, USA, and EU.

Probably the most significant development is that in recent EU Parliament's changes to the proposed Artificial Intelligence Act, the use of technology for biometric surveillance is prohibited except for law enforcement purposes provided that a judge should authorize the use of the latter; also the use of generative systems to indicate AI-created content.

Concerns and issues in the legal sector imply risk and mathematical difficulty; as such systems incorporate significant amounts of data, data privacy acts can be a problem. The bias can influence the results during training that may replicate social and historical discrimination based on race, caste, gender, and the kind of ideology as the models are trained similarly. Unlike trained attorneys such systems do not need one to be an attorney of the court and are not regulated by ethical courts' codes if they offer wrong or misleading legal information.

Predictive risk concerning the use of technology by judicial systems involves the over-reliance of judicial systems in technology-based recommendations may include an element of automated bias. For instance, there is a case where a New York lawyer was utilizing a generative tool for legal research where general cases had to be cited and he got sanctions and fines when the cases he cited, were not verifiable. This goes to show that there is a need for one to be careful when using these tools for legal research.

Further, the self-learning feature of advanced systems might create a technological and economic divergence, thereby threatening the architecture created by the Competition Act, of 2000. Therefore, the responsibility for technology-related mistakes that occur in the legal profession remains problematic, often impacting the existence and freedom of citizens. To deliberately avoid such situations lawmakers and industry gurus must actively drive the process to mark clear lines of culpability in the deployment of newer technology.