Hone raises $60M seed round at $285M valuation for business AI agents
Former OpenAI and Cognition staffers behind Hone secure major funding to build AI agents that autonomously manage business operations.
Key points
- Hone raised $60 million in seed funding.
- The company achieved a $285 million valuation.
- Founders include former employees from OpenAI and Cognition.
Hone, an artificial intelligence startup focused on autonomous business agents, has secured $60 million in seed funding. The round values the company at $285 million, according to a report by Bloomberg.
The investment highlights continued investor interest in AI applications that move beyond simple chat interfaces to execute complex business tasks. Hone was founded by a team of former employees from prominent AI firms, specifically OpenAI and Cognition AI Inc. The company’s stated mission is to develop AI agents capable of helping to run a business, implying a level of autonomy and integration deeper than current consumer-facing tools.
Context and Background
The founding team’s background at OpenAI and Cognition suggests a focus on advanced reasoning and autonomous action capabilities. Cognition AI is known for Devin, an AI software engineer, while OpenAI is the creator of ChatGPT and other large language models. By combining talent from these ecosystems, Hone aims to build agents that can navigate and operate within business environments.
Bloomberg reported that the group consists of former employees from some of the fastest-growing artificial intelligence firms. This pedigree is likely a key factor in the high valuation achieved during the seed stage, which is typically an early phase of company financing. The $285 million valuation places Hone among the higher-valued early-stage AI startups, reflecting the market's premium on teams with proven experience in building scalable AI systems.
Who is Affected
Businesses seeking to automate operational workflows are the primary target for Hone’s technology. Unlike general-purpose AI assistants that provide information or draft text, Hone’s agents are designed to "run a business." This suggests applications in supply chain management, customer service automation, financial reconciliation, and other back-office functions that require sequential decision-making.
The rise of such specialized AI agents may impact the broader AI software market. Competitors offering robotic process automation (RPA) or traditional business process management tools may face increased pressure to integrate more advanced AI reasoning capabilities. Additionally, the influx of capital into this specific niche indicates that venture capitalists are betting on the viability of autonomous agents as a distinct product category separate from foundational model development.
What should you do?
Security and technology professionals should monitor Hone’s developments as the company begins to deploy its agents in real-world business environments. The shift toward autonomous agents introduces new risk vectors, including the potential for agents to make unauthorized financial transactions or access sensitive data without human oversight.
Organizations evaluating AI tools for business operations should assess how Hone’s agents handle data privacy and security controls. Since no specific technical vulnerabilities or CVEs have been reported for Hone at this stage, the focus should be on architectural security and access management. Professionals should also track further announcements from Hone regarding partnerships or product launches, which will provide more detail on the scope and capabilities of their AI agents.
Background: Large language models
Large language models are a specific type of neural network trained on text. Generative AI is the broader umbrella term for any system that creates new content, including images and code. Not all generative AI uses large language models, but most modern text generators do.
What to do and how to stay safe: Hone
- Sudden role changes in output signal a successful override of system instructions.
- Hidden metadata in images or documents often carries invisible injection payloads.
- Immediate isolation of the affected session prevents lateral movement within the AI workflow.
Prompt injections exploit the model’s ability to follow instructions by hiding malicious commands in input data. Isolate affected sessions immediately and implement external guardrails to filter inputs before they reach the model.
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
How much funding did Hone raise?
Hone raised $60 million in seed funding.
What is Hone's current valuation?
The company has a valuation of $285 million.
Who founded Hone?
Hone was founded by former staffers from OpenAI and Cognition AI Inc.
