The digitalization of the justice system is gaining momentum. But when it comes to artificial intelligence, there is a gap between expectations and statistical certainty. There is increasing criticism in judicial circles that algorithmic text generators are flooding the courts with automated briefs. However, the federal government lacks evidence for this. In her response to a query from the AfD parliamentary group, she explains that she has no concrete information about a current increase in AI-generated applications. There were no systematic surveys or reliable figures.
Read more after the ad
It is known that AI systems are also used in everyday legal practice to formulate pleadings, writes the lead Federal Ministry of Justice. However, this is not a completely new development.
At the same time, the state is pushing forward the use of AI on the other side of the judge’s table. There are numerous projects in the German judicial landscape in which algorithms have been tested or, in some cases, already put into real operation. According to the information, the aim is to relieve the burden on the judiciary in dealing with its mountain of cases and to optimize processes. The focus is on automating uniform, standardized work steps. This primarily included the automated anonymization of judgments, the structuring of procedural documents and the evaluation and extraction of relevant data.
Practical test in the federal states
Various country projects show what this looks like in practice. Relevant AI applications for anonymization are running under the names “Aleks” in Lower Saxony and Bavaria and “Jano” in Hesse and Baden-Württemberg. The latter went into regular operation in the civil courts in December.
Experience so far is said to be positive, as the algorithms have already achieved higher hit rates in routine tasks than manual processing. Another example is the “Maki” project in Lower Saxony. It is designed to make it easier to deal with mass proceedings and to support the legal analysis of files and pleadings.
A current case shows that concerns about unchecked AI content in legal transactions are not unfounded. Invented file numbers and freely hallucinated judgments in appeal briefs caused trouble. Berlin judges were therefore forced to reprimand the legal profession.
European guard rails
Read more after the ad
The federal government is aware of the risks associated with AI-supported decisions. It warns against distortions caused by incorrect or incomplete data, algorithmic deceptions and the human tendency to blindly trust automated suggestions. Lack of traceability and susceptibility to manipulation are also an issue. To counter these dangers, regulatory guardrails came into effect.
The use of AI in judicial and public prosecutor’s procedures is largely based on the EU’s AI regulation, according to the justice department. This consistently classifies such decision support systems as high-risk AI. This was accompanied by legal obligations relating to risk management, data quality, complete documentation, cybersecurity and human supervision.
Last year, the federal and state governments adopted a joint AI strategy for the judiciary to ensure legally secure and uniform application. Standards for risk management systems and binding guidelines on data governance and quality are currently being developed within this framework.
Human responsibility
There is a dogma of the German Basic Law that stands above technological developments: the constitutional framework of judicial independence sets limits to automation. The ministry emphasizes that judicial power is reserved exclusively for elected judges. The final decision in a legal dispute must always be made by a person and be personally responsible. AI should only act in a supportive manner. Human supervision serves as a control mechanism to detect and correct machine undesirable developments in a timely manner.
To ensure that judicial staff are able to perform this function, the executive relies on targeted qualifications. Competencies on technical, legal and ethical issues surrounding technology should be strengthened through new training and further education offerings. Part of the national AI strategy is to create a competency framework to establish a common understanding and derive tailored training measures.
Millions for the digital pillar of the rule of law
The change is being financed and driven forward through cooperation between the federal and state governments. The strategy adopted in the summer of 2025 forms the foundation, which, according to the current planning status, should be largely implemented by the end of 2026. As part of its digitalization initiative, the federal government contributes financially and technically to the states’ AI projects. This start-up funding also runs until December.
The new pact for the rule of law should then follow seamlessly. Its so-called digital pillar provides for further funding with a term of three years. The executive branch is still examining whether and to what extent there is a need for additional legislative action in the area of European data protection law or on liability issues relating to automated processing.
(hze)
