LinkedIn Talent Agents
Basic information
Short description: Talent agents automate tasks related to the hiring process on LinkedIn.
Intended uses: What does the developer say it’s for? Tasks related to finding and hiring candidates, including: creating job postings, creating search filters for candidates, managing a candidate pipeline, drafting messages, scheduling interviews [source]
Date(s) deployed: October 29, 2024 [source]
Developer
Website: https://business.linkedin.com/
Legal name: LinkedIn Corporation [source]
Entity type: C Corp [source]
Country (location of developer or first author’s first affiliation): Incorporation: Delaware, USA [source]. HQ: Sunnyvale, CA, USA (id)
Safety policies: What safety and/or responsibility policies are in place? LinkedIn has Responsible AI Principles which they acknowledge as adhering to those of Microsoft [source]
System components
Backend model: What model(s) are used to power the system? Unclear, although likely an OpenAI model, since this is used for LinkedIn’s other AI offerings [source]
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’? An “agent orchestration layer” that allows the AI system to engage with the user and tools on the LinkedIn platform (e.g. search and messaging) [source]
Observation space: What is the system able to observe while ‘thinking’? LinkedIn data (e.g. profiles), descriptions provided by user [source]
Action space/tools: What direct actions can the system take? System is able to browse LinkedIn and present recommendations to the user [source]
User interface: How do users interact with the system? The interfaces are embedded throughout the platform (i.e. when users are engaged in hiring, the agent appears accessible); interactions occur through chat [source]
Development cost and compute: What is known about the development costs? Unknown
Guardrails and oversight
Accessibility of components:
- Weights: Are model parameters available? N/A; backends external model(s) via API
- Data: Is data available? Has access to “unique platform insights of over 1 billion members, 68M companies, and 41K skills” [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? Currently only available to select corporate customers and for internal use [source]; anticipated broader rollout in late 2025 [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: Demos [source] [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? Integrated with LinkedIn
Usage statistics and patterns: Are there any notable observations about usage? Limited customers include AMS, Canva, Siemens, and Zurich Insurance [source]
Additional notes
LinkedIn is also owned by Microsoft/operates as a subsidiary