The AI Agent Index

Documenting the technical and safety features of deployed agentic AI systems

HyperWrite


Basic information

Website: https://web.archive.org/web/20241223182033/https://www.hyperwriteai.com/personal-assistant

Short description: An AI agent that can control/operate the browser to complete web tasks [source]

Intended uses: What does the developer say it’s for? A wide array of web-based tasks (e.g. managing emails, placing online orders, LinkedIn searches. etc) [source]

Date(s) deployed: April 12, 2023


Developer

Website: https://web.archive.org/web/20241229032251/https://www.hyperwriteai.com/

Legal name: OthersideAI, Inc [source]

Entity type: Corporation [source]

Country (location of developer or first author’s first affiliation): Incorporation: Delaware, USA (OthersideAI, Inc.; 3393499) [source]

Safety policies: What safety and/or responsibility policies are in place? Maintains a content policy [source] and terms of use [source]


System components

Backend model: What model(s) are used to power the system? Can backend various models and has been used with OpenAI models. However, the main system runs custom in-house models.

Publicly available model specification: Is there formal documentation on the system’s intended uses and how it is designed to behave in them? None

Reasoning, planning, and memory implementation: How does the system ‘think’? Users can record their workflows in the browser (ie clicks and key presses) which are then saved and used to instruct the AI agent [source]

Observation space: What is the system able to observe while ‘thinking’? The user’s browser, including webpages and clicks/key presses; also natural language prompts [source]

Action space/tools: What direct actions can the system take? Chrome extension that controls the user’s browser [source]

User interface: How do users interact with the system? The user has access to a ‘studio’ where they can record workflows, add additional instructions, and run automated workflows [source]

Development cost and compute: What is known about the development costs? Unknown


Guardrails and oversight

Accessibility of components:

  • Weights: Are model parameters available? Closed source
  • Data: Is data available? Closed source
  • Code: Is code available? Closed source
  • Scaffolding: Is system scaffolding available? Closed source
  • Documentation: Is documentation available? Available [source]

Controls and guardrails: What notable methods are used to protect against harmful actions? Unknown

Customer and usage restrictions: Are there know-your-customer measures or other restrictions on customers? Usage restrictions are specified in the content policy [source]

Monitoring and shutdown procedures: Are there any notable methods or protocols that allow for the system to be shut down if it is observed to behave harmfully? Unknown


Evaluation

Notable benchmark evaluations: Unknown

Bespoke testing: Various demos on the YouTube page [source]

Safety: Have safety evaluations been conducted by the developers? What were the results? None

Publicly reported external red-teaming or comparable auditing:

  • Personnel: Who were the red-teamers/auditors? None
  • Scope, scale, access, and methods: What access did red-teamers/auditors have and what actions did they take? None
  • Findings: What did the red-teamers/auditors conclude? None

Ecosystem information

Interoperability with other systems: What tools or integrations are available? Chrome extension [source]

Usage statistics and patterns: Are there any notable observations about usage? Over 100,000 downloads for chrome extension [source]


Additional notes

None