The initial hype has passed: artificial intelligence has become an integral part of many people’s digital daily lives.
With generative AI applications such as ChatGPT, the technology experienced a new surge in development at the end of 2022. Today, AI has finally reached the mass market. Users no longer access it solely via individual websites. Instead, they use AI directly within search engines, office applications or ERP software.
This development is being driven in particular by the major technology providers:
However, artificial intelligence is no longer just a topic for end users. More and more companies are making targeted use of AI to optimise processes and gain a competitive edge. This is also the case with ERP systems such as SAP S/4HANA.
But what exactly does the term ‘artificial intelligence’ mean? Since when have AI applications been available, and why are they becoming increasingly popular? What can AI already achieve today, and where are its limitations?
What do terms such as ‘machine learning’ and ‘deep learning’ mean? How is artificial intelligence changing both businesses and society in general? And how is SAP approaching AI?
On this page, you will find answers to these and many other questions relating to the field of artificial intelligence.
Artificial Intelligence (AI) is a discipline within computer science. In Germany, the synonym ‘Künstliche Intelligenz’ (KI) is also used.
AI deals with methods that enable machines (computers) to solve tasks in the same way that a human would using their intelligence.
Artificial Intelligence therefore encompasses not only aspects of information technology, but also those of psychology, neuroscience, linguistics, communication studies, mathematics and philosophy.
Computer science is therefore more of a means to an end. It brings the various disciplines together and enables their implementation.
In order for AI to solve tasks independently, it must be trained. To this end, experts use special algorithms and provide the AI with training data. Through continuous training, the AI improves steadily – until it can solve the tasks set for it on its own.
There are essentially two or three types of Artificial Intelligence:
1. Autonomous and automatic task completion (weak AI)
The most widely used form of AI today is ‘weak AI’, or weak artificial intelligence.
Solutions such as intelligent software assistants, internet search engines, self-driving cars and speech recognition systems already utilise AI. As a result, they are becoming an increasingly common part of our everyday lives.
Although such systems have already achieved an enormous level of performance, they are referred to in technical terminology as weak AI.
The reason for this is that weak AI solutions operate at a relatively superficial level of intelligence. They do not develop a deep understanding of problem-solving.
2. Mimicking human thought and behaviour (strong AI)
Strong AI aims to attain or surpass human intellectual capabilities.
Systems with strong AI can recognise patterns and learn. Above all, they can apply the knowledge they have acquired to many new tasks. This works even when these tasks are not covered by existing algorithms.
Strong AI acts actively and flexibly, on an equal footing with humans. However, systems with strong AI do not yet exist.
3. Artificial superintelligence
Systems with ‘artificial superintelligence’ theoretically possess consciousness and human characteristics such as emotions.
To date, it has not yet been possible to develop such superintelligence. There is heated debate amongst researchers as to whether the development of such AI is even feasible.
Artificial Intelligence, Machine Learning and Deep Learning are distinct concepts.
Let’s take a closer look at these approaches below to further highlight the differences.
Machine learning refers to mathematical methods that enable a machine to generate knowledge independently from empirical data.
Or, to put it more simply: machine learning allows computers to perform useful tasks without having to be programmed to do so beforehand. Only the algorithm is programmed.
From a technical perspective, machine learning therefore involves the use of algorithms (computational rules) that can independently recognise patterns and regularities in data. To do this, they must first be fed extensive data and trained.
During the development phase, a programmer ensures that the machine learning model is continuously adapted and optimised.
In this way, the algorithm becomes ‘smarter’ from one dataset to the next through pattern recognition. Ultimately, it can perform its task independently using new and unknown data.
The main objectives of machine learning are to
In the business environment, machine learning applications have the potential to relieve staff of tedious, unproductive tasks. This frees up resources for new areas and makes work more efficient and cost-effective.
For example, machine learning software can independently scan paper documents, recognise the text, initiate further steps and organise archiving.
A more complex scenario in which machine learning is already being used today is predictive maintenance. The algorithms employed are capable of detecting potential damage to technical equipment and fault patterns at an early stage. If necessary, they can then request maintenance as the next step.
Speech recognition on mobile phones, spam filters in email inboxes and even facial recognition in photo management are largely controlled by machine learning algorithms.
Even before the AI hype triggered by chatbots such as ChatGPT, many people had already come into contact with machine learning without realising it – for example, when they were shown personalised adverts.
Deep learning is a subfield of machine learning. It is a specialised method that utilises so-called ‘deep’ artificial neural networks (DNNs) and vast amounts of data to learn particularly efficiently.
The way it works is modelled on learning processes in the human brain. Based on available information, such systems can repeatedly link what they have learnt with new content, thereby continuously improving their knowledge.
At a certain point, the machine is then able to make predictions, take decisions independently and question these decisions. If the outcome of the decision is unsatisfactory, it is adjusted in a new attempt.
Humans do not normally intervene in this learning process. This is also the key difference from machine learning.
Deep learning is particularly well-suited to scenarios in which large volumes of data need to be analysed for models and patterns. Examples of applications here include speech, object and facial recognition.
In speech recognition, for example, deep learning enables systems to independently expand their vocabulary with new words and word variants. Other areas of application include autonomous vehicles and robots, AI in computer games, and predicting customer behaviour within CRM solutions.
Since its inception, artificial intelligence has made enormous strides. This is particularly true of recent years.
Many corporations are currently trialling applications of AI. The use of AI opens up opportunities to optimise their own business operations and even to reshape entire industries. But how did artificial intelligence come about, and what level of sophistication have such systems actually reached to date?
The origins of artificial intelligence date back to the 1950s. The term was first used as early as that decade at a scientific conference in the USA. The scientist Marvin Minsky is regarded as one of the founding fathers of AI.
In 1966, Minsky defined AI as follows: “Artificial intelligence is the science of making machines do things that would require intelligence if done by men.” Put simply, this means that artificial intelligence exists when machines perform tasks that would require human intelligence to carry out.
Another early milestone in AI was the Turing Test. It was developed in the 1950s by the British mathematician Alan Turing. The test was designed to enable a human to communicate via chat software in real time with other humans and a machine.
In the 1960s, the so-called ‘General Problem Solver’ was unveiled. This was an AI system capable of solving simple problems. In the same decade, the ELIZA software caused a stir. At the time, this chat system made it possible to simulate therapeutic conversations.
In 1996, a computer defeated the then reigning world chess champion, Garry Kasparov. In the decades that followed, the capabilities of artificial intelligence improved continuously. This was largely due to ever-improving storage capacity and computing power.
In 2011, IBM unveiled its Watson software. It was able to win the quiz show Jeopardy! against human opponents. AlphaGo is another AI programme that achieved what was previously considered virtually impossible: defeating the world’s best professional Go player in 2016.
Artificial intelligence is now firmly established in both our private and professional lives.
A key area of application is AI-powered assistance systems. Applications such as Siri and Alexa have played a major role in making AI accessible to a broad user base. At the same time, the latest AI technologies are increasingly enabling real-time translation, thereby facilitating communication across language barriers.
Furthermore, generative AI applications and chatbots are rapidly gaining in importance. Companies are using them specifically to create content and make communication with customers more efficient. These are systems based on large language models. They are capable of understanding and processing natural language and generating content independently.
Compared to earlier chatbots, these AI solutions can conduct significantly more complex dialogues. Areas of application include customer service, internal knowledge platforms and even initial medical consultations.
Another key area of application for artificial intelligence is predictive analytics. AI algorithms analyse large volumes of data from various sources to produce well-founded forecasts. This enables early identification of trends in customer behaviour, business figures or operational processes. In the financial sector in particular, AI is used for fraud detection, risk assessment and claims management.
Artificial intelligence is also playing an increasingly important role in the field of marketing. AI enables data-driven planning and optimisation of campaigns. Companies can predict more accurately which customers are most likely to respond to an offer, when and via which channel.
AI is also well established in service management. Here, for example, artificial intelligence handles the automatic classification of service requests. Furthermore, it assists with support interactions and provides relevant information to help resolve issues.
Artificial Intelligence is also driving significant progress in IT security. Modern AI systems detect attack patterns at an early stage, identify anomalies and prioritise security incidents. Through machine learning, they continuously improve by analysing new threats and learning from them.
SAP’s strategic direction clearly shows that the company is increasingly evolving into a platform provider for AI-powered business processes. SAP is pursuing what is known as an ‘AI-native’ approach, in which AI is no longer used only in isolated instances, but is an integral part of all applications.
SAP aims to use artificial intelligence not only to automate business processes, but also to optimise them in a context-based manner. The focus is on seamless end-to-end processes.
Since 2023, SAP has been consistently driving forward the integration of AI. The company has grouped the relevant functions under the term SAP Business AI. At the same time, the portfolio is being systematically expanded to make AI usable across as many business areas as possible.
As part of this development, SAP is increasingly focusing on generative artificial intelligence and so-called ‘Agentic AI’. These AI systems are capable not only of generating content, but also of independently taking on tasks within business processes and coordinating multiple steps.
An important milestone in this development was presented at SAP Sapphire Orlando 2026: the vision of the Autonomous Enterprise:
The AI assistant Joule plays a central role here, serving as a unified user interface for SAP applications. Users interact with the systems using natural language to retrieve information, initiate processes or carry out analyses. In doing so, Joule utilises the business context to deliver precise and relevant results.
Behind the scenes, specialised AI agents are deployed to support or automatically execute entire business processes – for example, in finance, the supply chain or human resources. The aim is to achieve seamless end-to-end automation across all areas of the organisation.
Artificial intelligence brings with it enormous societal changes and challenges relating to data protection. SAP has already examined these ‘downsides’ of AI in depth.
The software group has therefore developed guidelines designed to govern the introduction and development of AI components. The overarching aim here is to ‘improve the functioning of the global economy and people’s lives’.
The principles cover the following aspects:
SAP also aims to address the societal challenges arising from AI. Aspects such as economic redistribution, economic development, social security and normative issues play a role here.
Artificial intelligence gives rise to a number of ethical dilemmas.
Artificial Intelligence is undoubtedly a highly emotional topic. Extreme positions often dominate the discussion.
One camp sees AI as a threat to all of humanity, while the other often sees the technology as a panacea for all of our problems.
No one can yet judge whether one of these scenarios will come to pass. The fact is, however, that AI will become significantly more important in the coming years and decades. Experts agree that Artificial Intelligence is a key technology of the digital revolution.
For companies that want to make progress in terms of digitization, there is therefore hardly any way around using the possibilities of Artificial Intelligence for themselves and driving forward automation via AI. SAP already offers a wealth of opportunities for this.
And where does that leave humans? Will they soon be completely replaced in the working world by robots with Artificial Intelligence? Certain job profiles could indeed disappear as a result of AI, but at the same time new jobs will be created.
Human skills such as creativity and empathy could once again come to the fore. There will be freedom to concentrate on one's own strengths again and to develop innovations. This also has a positive effect on employee satisfaction.
The intuition factor will also continue to be in demand. Decisions will be based much more on data from intelligent analyses. However, if the decision-making scope is high, the human being will still have the last word for the time being.
There is a chance that AI will develop in a positive way, empowering people to better solve the problems of modern society and achieve more.
One risk is allowing AI to operate beyond the bounds of sensible control. For businesses, such an approach would be disastrous, not only in terms of reputation and ethics. The level of security and control in the use of artificial intelligence could therefore determine whether intelligent machines become a curse or a blessing for us.
