We plan to connect the best researchers and developers from various countries on the field of AI & Law with investors who believe that well planed and produced modules of AI could be of great value for different future (say - 5 year) projects involving e-justice, e-administration, e-forms, e-legislation, online created documents, etc. The newly established company will try to take and maintain the leading position in this area of endless opportunities. We truly believe that this practical collaborative approach would first of all raise general awareness about tangible benefits AI would bring into the legal world, and secondly, motivate investors to help with the current and future projects and research.
In the past few decades there has been a lot of activities and research in the field of Artificial Intelligence (AI) in general, as well as in the area of Law. A lot of money has been spent, all over the world, many associations have been established, thousands of researchers were involved, a lot of conferences and other events have been held, countless articles have been published… but all that was mostly on the academic level. Industry was not (yet) interested – the subject seemed too removed from the real business opportunities. AI was left to science fiction for the time being…
Not any more! On one side – many tangible results have been achieved and quite a number of techniques evolved to a near mature stage. On the other side – many real-life processes evolved sufficiently enough to need badly the practical benefits of the AI in Law. Hence, significant business opportunities are showing up…
That is why we consider it is now time for an intensive commercial approach in this area and on the global scale! The initiative is still in the very early phase and the further development needed is so demanding that this could only be a medium (if not long) term investment. Our estimation is that we would need at least 5 years to develop and wrap up some very good and useful (commercial) modules (packages). At the same time, the results could be extremely beneficial and of high market value, provided that they are well guarded (with patents and other means of intellectual property protection). Furthermore – it is by the nature of the subject, that it would not be easy for any potential competitor to “steal” the results of such a serious and complex enterprise that will always have to be backed and (permanently) supported by state of the art development teams (the same that have produced the original product).
We need to realize that the potential end-users (buyers) of the abovementioned AI in Law modules (packages) will be mainly very big entities with huge budgets, who will integrate them with their own business processes. Governments already are and will be investing even more in their court systems in order to modernize them and make the judicial process more transparent and efficient. If there will be a module on the market which would help judges deliver their sentences faster and with more accuracy, they would likely implement it. Every large project of e-government and e-administration (with user friendly interfaces and involving “real life situations”) will include some kind of a legal expert system. The public administrators will not only accept applications from the citizens online, but will increasingly produce instant responses. More and more legal (and other) documents will be produced online and even various competitive companies will be interested to implement an abstract module of AI if it will give good results. Even the existing leading (legal) information providers will not hesitate to add enriched functionality to their specialized search engines.
Having a very good understanding with both – researches and developers from around the world as well as understanding of the real practical needs of many potential users, we plan to establish a new commercial company with the following goal – to be the leading high-tech provider of the artificial intelligence modules*, ready to be adapted and implemented within various other projects in the area of law. The company would most probably be named– ‘AI in Law, Future Technologies’. Because of the better patent protection possibilities, the company or one of its affiliates would be registered in the USA. Minimum starting capital will be about 5 MIO EUR.
There are various possibilities for attracting investors such as via direct invitations, single big investor, a few starting partners, joining VC later, IPO, etc. In all these cases the investor(s) will have to accept the idea, believe in the project and have full confidence in people who started this initiative. It is possible that some partial “by-products” will be ready for early market exploitation but on the whole the development will be focused on achieving original, well planned and documented, very demanding goals, within a 5 year schedule. The return on investment will therefore be delayed, but the financial expectations of the investors will be very high.
We plan to invite all the best developers and institutions from the field of the AI in Law (that have already achieved remarkable results) to join us or to cooperate with us in all possible ways. But it is quite clear already that the starting team and the institutions involved will be able to demonstrate from the very beginning many distinguished achievements and provide excellent references including the following:
Environment and Tools for Creating Legal Documents Online
Complex legal documents can now be created online, using special smart templates (procedures, questionnaires, instructions, comments, links, legal background etc.). Special portal for knowledge distribution is the place where authors and users meet in a friendly environment, using all the necessary tools via their browser. In the future, many more elements of AI will be added to the ‘smart forms’ thus elevating them to the ‘intelligent forms’ level.
State of the art legal information system
More than a decade of accumulated know-how and successful operation.
Specializing in development and integration of legal and business information systems. This includes content, knowledge and document management, processing and publishing over the Internet.
XML driven Content Management System with “Timeline” functionality
Main features include automatic recognition of the content and structure of non-structured data, versioning with full change history, cross linking and referencing of various data sources, granular customizable security and point in time functionality which provides date based searches
XML specific environment for legislative drafting and handling
Fragmentation of legislative information and inconsistencies of different legal document formats represent historical obstacles to a systematic organization of a normative corpus.
In today’s information society, the spread of legal information at the national and trans-national levels calls for an improvement of the overall quality of legal documents generated within different legal systems. For these purposes standardization in legal information is of paramount importance. It allows:
- to guarantee interoperability among
different legal information systems;
- to enable a straightforward achievement of integrated services for both policy makers and citizens;
- to enable communication among legal practitioners and towards European citizens, to make them aware and involved into the policy cycle and the decision making process;
- to allow Public Administrations, as the main providers and distributors of legal information, to distribute and sell reliable, updated and high quality public information, and therefore decrease the cost of public services;
- to upkeep legal orders in terms of coherency and consolidation;
- to improve the information market and to create a source of business for private companies.
In different European countries initiatives have been launched on the development of standards for legislation. Currently in Europe different national initiatives have introduced standards for legislation as the “NIR project” in Italy (www.nir.it), “LexDania” in Denmark, “MetaLex” in the Netherlands, CHLexML in Switzerland. Other similar initiatives in UK have established specific metadata models, such as the “Legal and Advice Sectors Metadata Scheme” (LAMS) that was developed by the Lord Chancellor’s Department as part of the Community Legal Service (CLS).
Such standards are mainly defined within the URN and XML frameworks which are respectively able to identify documents and describe their structure and semantics within a distributed environment. In order to make easier the adoption of standards for legislation, automatic or semi-automatic tools providing specific facilities are highly desirable. They should be aimed to handle legacy contents and to produce new legislative documents according to the defined standards.
Legacy legislative content management
Automatisms for legacy content handling represent key factors to promote the adoption of legislative document standards.
Modules can be implemented to convert legislative documents from their native formats (plain text, html, doc, etc.) to the XML format, as well as to detect cross-references and automatically construct the related URN.
Similarly, modules able to automatically or semi-automatically extract the semantics of legislative texts according to specific semantic models can be forseen. They should be able to support the intellectual activity to classify documents. The detection of semantics within legislative texts allows to provide citizens and legislators of advanced applications and services for legislative information systems: from semantic search facilities, to advanced access to norms, to consolidation, to monitoring the impact of new provisions on the legal order. In particular automatic consolidation can be achieved, so to obtain the updated versions of legislative texts, subjected to textual amendments and chronological versioning.
Such results can be achieved through the implementation of AI techniques; in particular machine learning techniques for document classification and information extraction can be implemented.
Editor for new legislative documents
XML standards for legislation usually cope with a wide and complex subset of documents: basically legal measures and regulative acts. The production of new documents, as well as the transformation of legacy contents according to the XML standards, can be a hard problem to face without an editing system guiding and supporting the user. Even though programs for XML drafting already exist, they have limits whether used for a specific class of documents, especially as concerns the generality and inadequacy of their editing functions with respect to the needs to implement specific XML legislative standards. Other possible general-purpose solutions, available on the market (Microsoft Word, Open Offce) suffer from the same limitations.
For these reasons a specific editor able to handle legislative documents in their native XML format, according to a specific XML standard, can be implemented. Further than integrating legacy content managers, such editor can be effectively used to produce new legislative documents at different steps of the legislative workflow, providing facilities to implement XML and URN standards, including metadata, as well as to follow the legislative iter of a new bill.
Specialists for machine learning
Laboratory for AI
All these areas are very well developed in Slovenia, mostly on the university and research institute level. We plan to invite everybody to join us.
There are still a few dilemmas that we are currently dealing with. We plan to resolve them as soon as possible taking into account the feedback and the interest shown for this project. Let’s just mention one…
Since we have noticed quite an interest among big commercial investors (including some from the USA and Canada) it would be possible to start the project on a truly international level. However, there are some indications that it could be better to seek support and commitment from the European Union – in which case the new company would be more EU oriented. It is therefore our intention to contact some of the highest EU officials and find out if there are any existing plans to make AI development a priority which would benefit from dedicated R&D funds. In that case the whole project could be designed around better collaboration and coordination with public authorities and institutions. We believe that our ambitious plans are complementary with the general objective of the EU which wants to become a competitive information society. In addition, it would probably be in the best interest of the Slovenian government to help this knowledge related project during its Presidency of the EU in 2008!
* Possible Modules to be developed
1. Court proceedings
2. On-line mediation
3. Administrative proceedings
4. Corporate compliance
5. Normative activities
6. Legal guidance
7. Modules for existing legal information systems
Ljubljana, 4th of April 2006
Anton Tomazic & Co.
(attributed by specialists from Slovenia, Italy, UK, Croatia, Canada, USA)
Illustration fom the Lab