Facing Enterprise Software Giants: OpenAI Seeks "Disruption," Anthropic Chooses "Symbiosis"
```
OpenAI and Anthropic are adopting sharply different strategic postures in the enterprise software market—one openly declares its intent to disrupt, while the other deliberately emphasizes partnerships. Analysts believe this divergence is reshaping market expectations for how AI will impact the traditional software industry.
On February 24, Anthropic released new feature details for its Claude Cowork AI software, demonstrating how enterprises can use this tool to access and invoke data stored in enterprise applications such as DocuSign, LegalZoom, and Salesforce.
After the announcement, enterprise software stocks that had previously been pressured by AI disruption expectations generally rebounded—Figma’s stock rose 10%, ServiceNow increased by 1.4%, Salesforce gained 4%. The positive market response indicates that Anthropic’s actual positioning is not as aggressive as some investors had previously feared.
Meanwhile, OpenAI last week made it clear at an investor meeting that its AI agents and future products will have the capability to replace software from tech companies such as Salesforce, Workday, Adobe, and Atlassian. It also presented revenue data from these enterprise software companies, contrasting them with its forecasted 2030 revenues.
Analysis points out that this private statement forms a sharp contrast with Anthropic's public posture and has further intensified the defensive stance of the enterprise software industry.
Anthropic bets on "replacing manpower" rather than "replacing software"
The tech media outlet The Information reported that when Anthropic launched its new Claude Cowork features, it intentionally focused its narrative on replacing manpower, rather than replacing software tools. The logic is:
New AI tools will continue to call upon existing enterprise software; enterprises still need to pay for these software tools, so the income sources for software vendors will not disappear.
During the launch event, Anthropic specifically invited its chief economist Peter McCrory. McCrory stated that the impact of AI on the labor market will be "highly uneven"—high-skilled workers will use AI to increase productivity, while low-skilled employees engaged in basic data entry tasks face the risk of being replaced.
PitchBook Data analyst Derek Hernandez further pointed out that Claude's features targeted at the financial services sector may put jobs in that industry under pressure, "the positions affected are those in investment banking, equity research, especially entry-level white-collar jobs."
The report notes that it's noteworthy that Anthropic’s internal employees are still using a variety of traditional enterprise application software. While this detail cannot entirely eliminate concerns among software vendors—after all, their business model depends on a large enterprise user base—it does convey a market signal very different from OpenAI's.
OpenAI plays the "disruption card" for investors
By contrast, OpenAI’s ambitions for the enterprise software market are much more overt.
According to reports, OpenAI last week explicitly listed Salesforce, Workday, Adobe, Atlassian, and other enterprise software giants as potential targets for replacement, and showed the revenue scale of these companies alongside its own 2030 revenue forecast to convey its vision of market potential to investors.
OpenAI also disclosed a calculation to investors:
An average employee using ChatGPT saves about 50 minutes of work time per day, equivalent to around $50 saved per person per day. Meanwhile, enterprise ChatGPT subscription starts at just $25 per employee per month. Based on this, OpenAI believes they are currently capturing only a small fraction of the value created. This calculation partly references data estimated by OpenAI shareholder Ark Invest.
The report states that these private statements corroborate the widespread interpretation of OpenAI's launch of its new "Frontier" AI product last month—OpenAI is attempting to position its technology as the top layer entry point for enterprise applications, seeking to control how enterprise data is accessed and gradually influence enterprise decisions about software and AI procurement.
Traditional software vendors build defenses; customer behavior is quietly changing
Facing the encroachment of AI labs, traditional enterprise software vendors are actively building defenses.
ServiceNow and Microsoft are emphasizing to customers that their software is more reliable and compliant compared to experimental products from AI labs. HubSpot is considering charging extra fees for customers who want to use AI agents to call data within their systems.
However, whether traditional software vendors can genuinely block the infiltration of AI agents remains uncertain. The core design logic for AI agents is to take control of users' computers and operate various applications in an automated, human-like way.
Analysts note that the most formidable moat for traditional enterprise software vendors is their long-term experience in global data compliance and privacy regulation—abilities that experimental AI products cannot replicate in the short term.
Although most enterprise clients have not yet truly attempted to replace existing enterprise software with AI, AI agents are changing how employees interact with software and are already creating substantial cost substitution effects in specific scenarios.
Reports state that a cybersecurity executive said he avoided more than $100,000 per year in CrowdStrike product subscription fees by connecting to AI agents. The CrowdStrike product was originally used to automate management of employee accounts, including flagging suspicious logins and identifying dormant accounts for suspected departed employees.
The executive switched to an AI agent provided by startup Torq—the agent, powered by OpenAI and Anthropic models, directly connects to raw login data gathered by Microsoft software within the enterprise, achieving equivalent functionality at a much lower cost.
The report says this case reveals the potential disruptive path AI may take in the enterprise software market: even if companies do not completely abandon existing software for now, AI agents may render some software products "optional" for employees, gradually establishing AI itself as the core tool for work. CrowdStrike stated that it has allowed AI agents to access its software to enable customers to use both together.
Risk warning and disclaimerThe market involves risks, and investment should be cautious. This article does not constitute personal investment advice and does not consider the particular investment objectives, financial situation or needs of individual users. Users should consider whether any opinions, views or conclusions in this article are suitable for their own circumstances. If you invest based on this, you are responsible for any resulting risks. ```