What is meant by the term AI?
Artificial Intelligence (AI) is not a new invention of the 21st century, but has been known since Alan Turing's question "Can machines think?" in the 1950s. Interest in AI is booming, especially in the 21st century, due to the ever more powerful computers and the voluminous flow of data.
The exponential increase in the immense volume of data in recent years has been of particular benefit to the AI technology. The International Data Corporation forecasts that 163 zettabytes of data will be created and replicated in 2025, ten times the amount of data in 2016. One zettabyte is a unit of measurement for storage capacity and stands for 1021 bytes. This is a trillion bytes, or in numbers 1,000,000,000,000,000,000,000 bytes. This in turn corresponds to 1,000 exabytes or one billion terabytes.
Before asking the question whether AI is defined as a robot superior to humans or as a helpful addition to daily activities, the approach of AI must be characterized. At the core of the capabilities and limitations of AI-based systems are algorithms, i.e. instructions broken down into individual steps that can be processed by a computer and mapped in software, as well as data processed by the system itself. In the context of AI systems, data act as a special component; they are not a pure substrate, but rather the data serve as a kind of building material for an AI system, since such a system primarily replicates behavior from data patterns fed into it.
A generally accepted definition of AI has - as far as can be seen - not yet been established.
From the indomitable chess player to self-propelled cars to a form resembling a human being: this is how the most diverse types of AI can be identified. One of the special features of AI is undoubtedly its autonomy, so a fundamental distinction with regard to the degree of autonomy seems to make sense. AI with a weaker degree of autonomy can be used selectively, whereas AI with a high degree of autonomy aims to create a general intelligence that is equal to or superior to human intelligence.
Legal Basis of AI:
Based on the definition in the European Commission's opinion on AI of 25 April 2018, AI can be described as a system that reflects intelligent behavior by analyzing the environment and - with a certain degree of autonomy - taking measures to achieve certain goals.
At present, the legal framework concerning AI is rare. There is hardly any jurisprudence that knows AI-related laws or regulations. Furthermore, there is uncertainty as to the extent to which laws are applicable to AI. Studies such as those conducted by Stanford University highlight the need for regulation (including legal regulation).
It is a true balancing act to make the best possible use of the advantages of AI and not to regulate it to the extent that the development is stopped, but still to guarantee the current legal standards, in the form of intact security and liability provisions or respect for private life. The safeguarding of or compliance with ethical principles is another challenge posed by AI technology.
The need for laws and standards is ultimately left to the respective countries; they have various options for standardization or non-regulation.
AI and Due Diligence:
Due diligence is of major importance in M&A transactions. In the field of company and investment acquisitions, this is generally understood to be the pre-buy examination of the target company with regard to legal and other aspects, which is usually carried out by the prospective buyer or its advisor (buyer due diligence, buyer or buy-side due diligence). The focus of these examinations is the determination of the value of the acquisition object and the opportunities and risks associated with the purchase of the company or investment..
International transactions are increasing in value and scope, and the review and analysis of information is becoming more extensive, also in legal terms. Nevertheless, from an entrepreneurial point of view, one tries to keep these costs as low as possible, but without taking on unrecognized risks. An AI system could therefore respond to the individual needs of the user and offer the possibility to optimize the performance of the due diligence process. Time-consuming thinning would thus be a thing of the past, allowing the focus to be shifted to critical facts and legal issues. The software would bring transaction costs down to a lower level and reduce the duration of the process.
AI is not science fiction, it is rather regarded as a compliance tool that brings new opportunities, but also challenges. The "arms race" for a technological lead has long since begun, not only for companies, but also for state authorities and governments. In the near future, it will become clear in which direction legislators want to steer AI technologies.
The European Union has set itself the task of preparing the entire population for the forthcoming challenges resulting from AI technologies. This process was initiated with the European Union's recommendations for civil regulations in connection with robots. The European Parliament also points out the importance of introducing regulations on legal and ethical aspects without stifling innovation. We can therefore expect to witness an increase in discussions related to AI regulations and its legal framework in the near future.
 Turing, Computing Machinery and Intelligence, https://www.csee.umbc.edu/courses/471/papers/turing.pdf (queried on 6. July. 2020).  Petsche/Heiss in Petsche/Mair (Hrsg), Handbook ComplianceAufl. 3 (2019) Artificial Intelligence and Compliance, page 461.  IDC, Data Age 2025, https://www.seagate.com/www-content/our-story/trends/files/Seagate-WP-DataAge2025-March-2017.pdf (queried on 6. July 2020). https://www.computerweekly.com/de/definition/Zettabyte#:~:text=Ein%20Zettabyte%20ist%20eine%20Ma%C3%9Feinheit,n%C3%A4chsth%C3%B6chste%20Einheit%20nur%20Yottabyte%20vorgesehen. (queried on 6.July.2020).  Von Bünau in Breidenbach/Glatz, Rechtshandbuch Legal Tech (2018).  Blues in Hartung/Blues/Halbleib, Legal Tech Rz 1165.  Artificial Intelligence for Europe, https://ec.europa.eu/digital-single-market/en/news/communication-artificial-intelligence-europe (queried on 6. July 2020).  Petsche/Heiss in Petsche/Mair (Hrsg), Handbuch ComplianceAufl. 3 (2019) Artificial Intelligence und Compliance, page 464.  Stanford University, Artificial Intelligence and Life in 2030, https://ai100.stanford.edu/sites/default/files/ai100report10032016fnl_singles.pdf (queried on 6 July.2020).  Petsche/Heiss in Petsche/Mair (Hrsg), Handbuch ComplianceAufl. 3 (2019) Artificial Intelligence und Compliance, page 465.  Petsche/Heiss in Petsche/Mair (Hrsg), Handbuch ComplianceAufl. 3 (2019) Artificial Intelligence und Compliance, Seite 465.  Knauder in Knauder/Marzi/Temmel (Hrsg), Handbuch Wirtschaftsverträge (2. Lfg 2013) Due Diligence, page 4.  Petsche/Heiss in Petsche/Mair (Hrsg), Handbuch ComplianceAufl. 3 (2019) Artificial Intelligence und Compliance, page 469.  Petsche/Heiss in Petsche/Mair (Hrsg), Handbuch ComplianceAufl. 3 (2019) Artificial Intelligence und Compliance, Seite 469.  Parlaments Civil Law Rules on Robotics, http://www.europarl.europa.eu/sides/getDoc.do?pubRef=-//EP//TEXT TA P8-TA-2017-0051 0 DOC XML V0//EN (queried on 6. July. 2020).