Artificial intelligence is currently the dominant topic in the technology sector. Initially, investors unanimously saw enormous growth potential in new AI applications. In the meantime, however, the risks are increasingly coming into focus. Investors are wondering whether software solutions themselves could be replaced by AI agents from new providers such as OpenAI. Some investors are therefore generally avoiding the software sector. For example, Softbank as a venture capitalist focuses on topics such as AI data centers, semiconductors and robotics when making new investments and explicitly avoids the software sector. The majority of shares in the software sector are trading at double-digit losses and are often cheap from a historical perspective.
It is quite possible that the market is exaggerating somewhat here. It is certainly difficult to assess further developments in the field of artificial intelligence, as progress is rapid. However, business processes are very complex, especially in larger companies, and cannot easily be taken over by an AI agent, even if it is trained over a longer period of time. Expectations in this regard were rather disappointed at the OpenAI Developer Day in October. The all-encompassing AI agent that can map the majority of business processes was not presented. The current trend is for OpenAI to establish itself as a kind of new operating system similar to Apple's iOS or Microsoft's Windows in the field of AI, on which existing software solutions are based. This has led to a recovery in software shares, at least on the day of the presentation.
Exclusive databases in the midst of AI disruption
Ultimately, however, it cannot be ruled out that at least some of today's software solutions will become obsolete in the long term and be taken over by AI agents. Investors should therefore focus on providers whose software solutions are not so easy to replace. Exclusive access to data is important in this context. A classic example is software solutions from the area of enterprise resource planning (ERP). The systems of the leading providers manage a large proportion of their customers' internal company data, which no other system can easily access. Accordingly, the risk of disruption here is likely to be rather low.
The leading German provider, for example, stores the data securely in its own database. AI therefore offers more opportunities than risks for the provider. In the medium term, it aims to generate one billion euros in AI sales. Although this would only be three percent of turnover, the CEO recently hinted that it could be more. The central tool is an AI agent which, once trained, can perform routine tasks in accounting, bookkeeping or human resources independently. The provider is also in the middle of cloud migration. A good half of the customer base has already migrated to the cloud solution. This year, however, the customs dispute has unsettled customers, particularly in the second quarter, meaning that the provider will only be able to achieve the lower end of cloud growth for the first time. It normally adjusts its outlook upwards over the course of the year. The continuing weakness of the US dollar is also having a negative impact. If the dollar were to stabilize, the situation would change. Tailwinds from cloud migration and sales potential from AI, SMEs and the new business platform (at least one billion euros in additional sales each) would also be possible. In such an optimistic scenario, double-digit sales and profit growth would probably be achievable by the end of the decade.
US software solutions in the battle with hyperscalers
The same applies to the leading US provider of corporate databases, which offers dedicated solutions for major customers and SMEs. In addition, it is now starting to build AI data centers on a large scale, which are to be rented out to third-party customers. OpenAI won a major contract worth USD 300 billion over five years (starting in 2027), which initially led to a rapid rise in the share price. The US provider has a total order backlog of USD 523 billion and outperforms the "established" major hyperscalers such as AWS (Amazon), Azure (Microsoft) and GCP (Google). The company expects its revenue to quadruple to USD 225 billion by 2025 and earnings per share to reach USD 21 by 2030.
The initial euphoria has now given way to fears as to how this expansion can be financed and whether OpenAI will even be able to meet its obligations given the still very high cash burn rate. The US provider's shares are currently trading at a price/earnings ratio of just nine based on the 2030 earnings target. The expansion of AI data centers is very capital-intensive and therefore poses risks for the US provider in particular, as the existing cash flows are not sufficient and part of it has to be financed through debt. However, the US provider emphasizes that it only wants to accept profitable orders and achieve a gross margin of 30 to 40 percent on the AI cloud business. If this is the case, the current market fears would be unfounded.
Complex software solutions: tend to be protected from disruption
The relatively complex software solutions in the area of product lifecycle management (PLM) also tend to have a low risk of disruption. Nevertheless, PLM providers were unable to escape the generally weak price trend in the software sector. The market leader in particular, which has a market share of 16%, has come under greater pressure as the outlook for 2025 had to be revised downwards. The life science business surrounding the specialist company acquired in 2019 is primarily revealing weaknesses in contract research organizations (planning, implementation, management and analysis of clinical trials for new drugs) and is currently only growing in the low single-digit range. In addition, the life cycle market leader has a high proportion of customers from the automotive industry, which has come under pressure. If the business can stabilize next year, it would have recovery potential, as the valuation for 2026, measured by the price/earnings ratio, is just under 18, well below the five-year average of 33.
Risks in image and video processing due to AI generation
Probably the greatest current AI disruption risk is seen in the leading provider of image editing software. There are a large number of new "text-to-video AI models" from providers such as OpenAI (Sora 2), Google (Veo 3) and Runway (N/L), which potentially represent new competition for the existing flagship products in image and video processing. Image and video processing are particularly well suited to AI. However, the majority of customers are likely to continue working with the market leader's established solutions, particularly in the professional sector. The market leader also offers its own AI tool. In addition, the current new AI solutions have not yet been able to score with such extensive functionalities. Furthermore, around 25 percent of sales come from the document sector (PDF), where there is no risk of disruption. Nevertheless, it cannot be ruled out that the market leader for image and video software will become an "AI victim" and, in the worst-case scenario, be completely replaced sooner or later. Accordingly, the share is only trading at a price/earnings ratio of 15 for 2026.
Possible risks for customer management solutions
The leading provider of customer relationship management systems (CRM) is also likely to face an above-average AI disruption risk. Although it also offers its own AI tool, monetization has fallen short of expectations, but with annual recurring revenue of USD 1.4 billion (corresponding to a good three percent of revenue), it is above average in a sector comparison. The provider is also increasingly focusing on AI data analysis. A provider of cloud data management was purchased for eight billion US dollars. This is now to be integrated into the existing solutions. The customer management provider itself also sees more opportunities than risks from AI. By 2030, the aim is to increase turnover organically by an average of ten percent to 60 billion US dollars. The company has also launched a USD 50 billion share buyback program (of which USD 28.1 billion has already been called up, USD 3.8 billion in the last quarter alone).
Conclusion: software is dead, long live software
Unfortunately, the AI disruption risk in the software sector is difficult to assess. However, the market could be exaggerating somewhat here, as it is questionable whether AI agents can really replace the majority of software solutions. Providers whose data is protected and not easily accessible, such as software solutions in enterprise resource planning (ERP), are likely to have a relatively low risk. Relatively complex software solutions, such as in product lifecycle management (PLM), are also likely to be difficult to replace. By contrast, sentiment is particularly negative for the PLM market leader (lowered outlook for 2025) and the US market leader for enterprise databases (financing the establishment of the new AI data center business), which recently led to sharp share price falls. The market leader in image and video processing and the market leader in customer management systems certainly have an increased risk of disruption. Both companies would have valuation potential if the risks currently priced in turn out to be exaggerated.
Legal information
Marketing advertisement - All information published here is for your information only and does not constitute investment advice or any other recommendation. The statements contained in this video reflect the current assessment of DJE Kapital AG. These may change at any time without prior notice. All statements made have been made with care in accordance with the state of knowledge at the time of preparation. However, no guarantee and no liability can be assumed for the accuracy and completeness. The video and its content are protected by copyright. Long-term experience and awards do not guarantee investment success. Securities are subject to market-related price fluctuations that may not be offset by the active management of the asset manager. Please note our risk warnings.