HumanX Conference Observation: AI intelligent entities' investments are soaring, UBS claims OpenAI and Anthropic are transforming into enterprise IT budget "devourers".

After participating in the HumanX Global AI Conference held last week in San Francisco, analysts from the international investment giant UBS Group AG stated that OpenAI and Anthropic are transforming from AI model supergiants to "budget eaters" of enterprise IT budgets. UBS Group AG analysts wrote that global cutting-edge AI companies like Anthropic and OpenAI launching more powerful AI models pose a significant threat to the fundamental growth prospects of most traditional software companies - these cutting-edge AI companies have strong enough capabilities to take a larger share of enterprise customers' wallets from them. Enterprise wallet share mainly refers to how much spending share a technology industry supplier takes from the total IT/software/digital budget of enterprise customers. According to the UBS Group AG analyst team, enterprise customers who used to allocate significant funds to traditional software vendors for enterprise-level IT budgets are now being taken more share by cutting-edge AI companies like OpenAI and Anthropic. Previously, enterprises may have spent the majority of their funds on SaaS applications, BI tools, customer service software, low-code platforms, and collaboration software; now, customers are more likely to build generative AI application solutions and AI agents directly using AI application platforms like Claude/ChatGPT. This portion of the budget may shift from traditional software companies to AI model vendors on a larger scale. The market's concerns and cautious stance on software stocks are not unfounded! At the HumanX AI conference held in San Francisco last week, about 6,000 investors, entrepreneurs, and tech company executives gathered. This high-profile conference not only outlined the deepening relationship between artificial intelligence technology and practitioners in various industries but also sparked intense debates about the impact of AI on employment. "One change we noticed at HumanX was the significant increase in enterprise AI investments, which have already expanded far beyond Microsoft Corporation's release of Copilot licenses and AI programming tools for large software engineering use cases," UBS Group AG analysts Karl Keirstead and Dean Marriott wrote in an investor report. "Ultimately, the latest enterprise customer research results support the pessimistic expectations already reflected in the pricing of these stocks and the more cautious outlook on SaaS/application software companies." Enterprises' urgent need to improve efficiency and reduce operating costs has greatly driven the widespread use of two core categories of AI application software - generative AI applications and AI agents. Among them, AI agents (i.e., AI agents that perform a variety of tedious and complex tasks independently) are likely to be the ultimate trend in AI applications in the next decade. The emergence of AI agents signifies that artificial intelligence is transitioning from an information assistance tool to a highly intelligent productivity tool. According to the latest research by MarketsandMarkets, the AI agent market is expected to reach as high as $53 billion by 2030, with a staggering annual compound growth rate (CAGR) of 46% starting from 2025. The explosive emergence of AI agents like Anthropic's Claude Cowork and OpenClaw in 2026 is not coincidental; fundamentally, it is the first convergence of five curves - "model synthesis capabilities, tool protocols, AI developer frameworks, reasoning costs, terminal contextual capabilities." At the AI application layer, AI agents are likely to become the predominant business interface, as they directly translate "intelligence" into "action," signifying the advancement of AI from "answering questions" to "execution, collaboration, and completion of extremely complex multi-step tasks." Anthropic's previous report on the 2026 agent coding trends explicitly predicts that the agentic workflow (i.e., agent-driven collaborative workflow) will expand from engineering teams to non-technical departments such as sales, legal, operations, and marketing. The core viewpoint of Anthropic is that agents are not "a specific function," but agents are becoming the default interaction paradigm for the next generation of software systems, functioning not only as models but as "context aggregators + action orchestrators," moving from Apps to Intent, and from tools to complete independent execution. OpenAI and Anthropic are transforming from model giants into "budget eaters" of enterprise IT budgets. According to information obtained from participants, OpenAI and Anthropic are becoming the "best-in-class AI application software companies." "Customers using Claude/ChatGPT to customize AI applications and focus on agent-driven workflows, as well as the trend of Anthropic/OpenAI launching first-party products, further increase the risks faced by existing software industry vendors in terms of future profit growth," they added. With a series of high-efficiency agent-focused AI products recently launched by leading AI model companies like Anthropic and OpenAI, it is highly likely that they will replace certain functional software services at a much lower cost, causing a heavy selling pressure on global software stocks. The pessimistic tone of "AI disrupting everything" since February is mainly due to the market's growing concern that agent-based workflows like Claude Cowork and OpenClaw (formerly known as Clawdbot, Moltbot) could weaken the entire software empire based on the subscription revenue model of SaaS seats, leading to rare sell-offs that quickly spread to industries such as cybersecurity, online education, traditional finance, insurance, real estate, transportation systems, and any other industries that appear to rely on seat revenue models or labor-intensive business models - the market believes that these industries will be completely disrupted by AI. Not only the US stock market, but the software sector of global stock markets has been continuously plummeting in the panic of "AI disrupting everything" since February. Although share buybacks in the US software sector have surged, investors are not buying in because the market is truly concerned about whether long-term fundamentals and business models will be completely reshaped by AI entities like Claude Cowork and OpenClaw. However, at least for the time being, some software companies that appear relatively safe in the wave of AI emergence are those deeply embedded in AI technology and focused on protecting and managing enterprise data, such as "AI + core operational processes" type software companies like Snowflake (SNOW.US), Palantir (PLTR.US), and Databricks. "Many research results have mentioned that data is a key advantage, and customers are also investing in modernizing and opening up their data - a trend we have been pointing out for a long time; therefore, we are very bullish on companies like Palantir, Databricks, and Snowflake," analyst Keirstead pointed out. "This data category is now under scrutiny due to AI risks, but apart from some people mentioning using Claude to replace the functionality of Tableau/Salesforce (CRM.US), we have not heard of any research suggesting that AI models like those under Claude/GPT will pose risks to data-centric software companies like Palantir, Databricks, and Snowflake." Regarding Microsoft Corporation, UBS Group AG states that market opinions are divided. "Regarding Microsoft Corporation, we have heard several opinions that Microsoft Corporation is 'lagging behind' and some disappointment with a series of innovative AI application products based on Copilot, but there have also been survey respondents who believe we should not underestimate it because Microsoft Corporation has very high market trust in enterprise security, access control, and comprehensive governance, and has the exclusive position to embed AI workflows deeply using this advantage," Keirstead pointed out. Furthermore, although Anthropic's Claude Mythos caused a sensation last week, it is still too early to determine the impact it may have on the subset market focusing on security software. The latest UBS Group AG research report shows that the market has not yet formed a well-founded judgment.