Competition is intensifying among leading AI model developers targeting enterprise AI agents. OpenAI and Anthropic are even setting up joint ventures with private equity firms to provide AI to portfolio companies of those firms.

OpenAI earlier this year unveiled its enterprise AI agent platform Frontier. Its strategy is to help companies create multiple "AI co-workers" and draw on data from various business applications. Anthropic, through Claude Code and Claude Work, is also helping enterprise users, whether developers or non-developers, use AI agents in real-world settings.

Big cloud companies led by Amazon Web Services and Microsoft are moving in the same direction. They are continuing to increase the funds they pour into building platforms that let businesses use AI agents safely and properly.

Cloud-based data platform companies such as Snowflake and Databricks have also recently been rolling out AI agent projects.

Palantir is not mentioned as often as OpenAI or Anthropic, but it is increasingly being recognised as a meaningful player in enterprise AI agents. Stability is critical for enterprise AI agents to work in practice, and Palantir is seen as having a lot to offer in that area.

Palantir does not develop LLMs like OpenAI or Anthropic. Palantir provides software that lets companies connect and organise data stored in existing systems such as Snowflake, Salesforce and SAP. It also supports building AI agents on that basis. Given that existing enterprise software companies are reluctant to open their data for external AI agents to use, this is a notable capability.

According to a recent report by The Information, Palantir is also drawing attention for its forward deployed engineer model, in addition to its software. Many people working in the enterprise software field acknowledge the advantages of this approach.

The core of the FDE approach is to stay on site at customers and implement Foundry, Palantir's ontology software, to fit internal circumstances. The Information said Salesforce, ServiceNow, Snowflake and even OpenAI view this model positively.

Instead of developing its own LLM, Palantir supports customers so they can choose suitable models, including open source, and use them for automated supply chain monitoring, workforce and inventory management, and sales data analysis.

Centrus Energy, which operates the only uranium enrichment facility in the United States, built a system by combining Palantir software with an Anthropic model to automate workforce deployment adjustments, supplier replacement and Nuclear Regulatory Commission report writing.

Palantir expanded into private companies after doing business with government agencies including defence. Government contracts still make up a large share, but the enterprise segment also grew explosively last year. Revenue from U.S. corporate customers rose 109% last year, and overall corporate revenue growth increased to 60%. That growth stands out compared with other enterprise software companies.

Palantir is also speeding up its push into the domestic market by partnering with major companies. After partnering with KT in 2025 to target South Korea's AI transformation market, it has recently teamed up with LG CNS as well.

Under a strategic partnership agreement, LG CNS will provide Palantir Foundry and AIP and other enterprise platforms in forms optimised for each customer. To do so, LG CNS will create a dedicated organisation for Palantir business called FDE. The FDE organisation will work with Palantir to identify and execute high value-added AI transformation projects across industries such as manufacturing, energy, electronics and logistics. It plans to move into full-scale business expansion starting with LG Group, which is actively considering adopting Palantir platforms.