
AI Devours Software! After the release of GPT-5, this week European and American software stocks have collapsed

The European software sector plummeted on Tuesday, with German software giant SAP's stock price briefly falling by 7.1%, resulting in a market value loss of nearly 22 billion euros, marking the largest single-day decline since the end of 2020. Several software stocks have accumulated double-digit losses since mid-July. After GPT-5, the market has first tangibly felt that the iteration of AI may not only enhance productivity tools but could fundamentally reshape the business models of the software industry
Concerns about artificial intelligence replacing traditional software finally triggered substantial panic selling in the market this week.
On Tuesday, the European software sector suffered a heavy blow, with German software giant SAP's stock price plummeting by 7.1% at one point, resulting in a market value loss of nearly 22 billion euros, marking the largest single-day decline since the end of 2020. French company Dassault Systèmes and the UK’s Sage Group also saw significant declines, with many software stocks having dropped double digits since mid-July.
This round of selling is closely related to the impact of the rapid iteration of AI models. Last week, OpenAI released the next-generation GPT-5, and a few days ago, Anthropic also launched a specialized version of Claude for financial services. Fund managers warned that the capabilities of the new models are sufficient to threaten the core businesses of some software and data service companies—including financial data providers, data analytics platforms, and even certain enterprise applications.
Kunal Kothari, a fund manager at Aviva Investors, pointed out that the Claude financial version directly impacts business models that rely on data provision, such as LSEG, stating, “The market is beginning to realize that each generation of GPT or Claude may have capabilities several times greater than the previous generation, which challenges the entire business logic.”
The "AI Devouring Software" Narrative Resurfaces
In fact, the notion of "AI devouring software" is not new. As early as 2017, Jensen Huang, CEO of NVIDIA, predicted that AI would change the landscape of the software industry. However, following the release of GPT-5, the reactions from Wall Street and European investors have been unprecedentedly real.
The analyst team at RBC Capital Markets stated, “The valuation of the software sector continues to be under pressure, and volatility will persist in the short term,” as the market is dominated by the narrative that “AI will lead to the demise of software.”
This sentiment has also spread to the U.S. market. On Monday, U.S. cloud-based collaboration platform Monday.com plummeted by 30%, while Salesforce and Adobe have seen declines of over 25%-30% this year, with a basket of software stocks relative to the semiconductor sector falling to its lowest point since January.
The core concern of the market is that as the capabilities of AI coding and generation applications improve, companies may build the tools they need themselves, thereby reducing their procurement spending on traditional "packaged software." Meta CEO Mark Zuckerberg even predicted earlier this year that by 2025, AI engineering agents with "mid-level engineer" capabilities could emerge
High Valuation Amplifies Impact
The recent decline in European software stocks has another amplifier—high valuations.
Currently, the average price-to-earnings ratio of the STOXX 600 is about 17 times, while SAP's price-to-earnings ratio is close to 45 times. UBS O’Connor Chief Investment Officer Bernie Ahkong pointed out, "High valuations make them exceptionally sensitive to any potential negative news."
However, some institutions believe that this wave of panic selling may create opportunities. Ahkong stated, "Some companies may ultimately turn AI into a profitable tailwind, but it will take time to prove."
Not All Software Will Be Replaced
Despite the market's loud slogan of "AI devouring software," some investors and analysts remind us that the software industry is not one-size-fits-all.
Steve Wreford from Lazard Asset Management believes that companies deeply embedded in customer workflows and possessing hard-to-replicate proprietary data still have strong moats. Paddy Flood from Schroders also pointed out that enterprise-level critical applications are less likely to be impacted due to high switching costs and the importance of service stability.
Aviva's Kothari cited the UK credit agency Experian as an example, stating that it not only possesses unique data but is also deeply embedded in the business processes of financial institutions, "You can't issue loans without it."
However, Kothari also warned that relying solely on data barriers may not be secure—AI could seize control of more segments in the data value chain.
Over the past half-century, companies like Microsoft and SAP have built the golden age of business software. But after GPT-5, the market is feeling for the first time that AI's iteration may not only enhance productivity tools but could fundamentally reshape the business models of the software industry.
As one fund manager said, " Maybe in 10 years, AI will really devour some software, but the market has already begun to price in this risk ahead of time."