Glean CEO Arvind Jain says company aims to be easily replaceable
Glean’s founder says the enterprise AI firm will store user data in standard formats to prevent vendor lock-in and avoid raising more capital.
Key points
- Glean has no plans to raise additional funding or pursue an immediate public listing.
- The company stores AI memory in standard formats so enterprises own their data.
- Glean acts as a context layer that can speed up other AI models like Claude.
Glean chief executive Arvind Jain stated that the enterprise artificial intelligence company intends to remain easily replaceable for its customers. According to a report by The Next Web, Jain emphasized that the firm will not rely on vendor lock-in strategies to retain business, distinguishing its approach from some competitors in the rapidly evolving AI sector.
Jain told The Next Web that Glean has no concrete plans to go public and no intention to raise more capital. While the company is preparing its internal processes for a potential listing, Jain noted that the AI industry moves too quickly for the stability required by public markets. He added that the company could list if it chose to do so in the future.
Business strategy and financial status
Glean reported passing $300 million in annual recurring revenue in May. This milestone came just 15 months after the company reached $100 million in annual recurring revenue. Jain founded Glean in early 2019 after spending more than a decade at Google as a distinguished engineer working on Search, Maps, and YouTube. He also co-founded data security company Rubrik and led its research and development efforts before starting Glean.
The company began as an internal search engine that connects to various corporate systems. These integrations include Salesforce, HubSpot, Jira, GitHub, and various human resources and IT tools. The platform learns where specific knowledge resides within an organization. Today, Glean operates as an AI coworker and an agent platform that utilizes models from multiple different labs.
Technical approach and data ownership
Jain explained that Glean prioritizes context over raw model power for most enterprise tasks. For 85% to 90% of tasks, the company prefers using models other than Anthropic’s Claude, which Jain described as very expensive. He noted that some GPT models cost an order of magnitude less while still handling the majority of that work effectively.
Glean builds knowledge graphs of a company’s people, concepts, and processes. Customers can use Glean as a context layer for other AI tools. When Claude requests context from Glean instead of searching 100 systems independently, Glean’s studies show the process becomes twice as fast. This method also halves token usage and makes the resulting work preferred five times more often.
Data portability and privacy controls
Jain stated that Glean does not want to win through lock-in mechanisms. As users interact with AI coworkers, those tools build memories of individual work preferences. With providers like Claude and OpenAI, the provider owns that memory. In contrast, Glean stores this information in a standard format that the customer company owns. The same applies to what agents learn from human feedback.
“Our goal is to actually make us very easily replaceable,” Jain said. Regarding privacy, Glean follows the permissions a company has already established. One employee cannot view another’s inbox through the platform. These checks are programmatic rather than left to AI, because AI can be fuzzy and prone to making mistakes.
What should you do?
Security and technology professionals should evaluate whether their enterprise AI tools store user interaction data in proprietary or standard formats. Organizations should verify that AI platforms respect existing identity and access management permissions rather than creating new data visibility paths. Teams should assess the cost implications of different model providers for routine tasks. Leaders should review vendor contracts to ensure they retain ownership of AI-generated memory and context data.
What to do and how to stay safe: Glean
- Review enterprise AI vendor contracts to confirm who owns the data generated by AI agents and user interactions.
- Verify that AI tools integrate with existing identity providers to enforce strict, programmatic access controls.
- Assess whether AI platforms offer data export options in standard formats to ensure portability.
- Monitor vendor financial health and public listing plans to gauge long-term product stability.
Step-by-step guide: Prompt Injection Warning Signs: How to Spot AI Manipulation
General security guidance from the Threat Times newsroom. It is not confirmed advice from the organisations named in this story.
Frequently asked questions
Does Glean plan to go public soon?
No, Glean has no concrete plans to go public and no plans to raise more money, according to CEO Arvind Jain.
How does Glean handle data ownership for AI memory?
Glean stores AI memory and agent learning in a standard format owned by the customer company, not the vendor.
Can employees see each other's data through Glean?
No, Glean follows existing company permissions, so one employee cannot see another’s inbox or private data.
