Auto-Apply Safety Policy

How JobCompanion ensures safe, transparent, and consent-first automation — with hard blocking rules, audit logging, and user approval at every step.

Last updated: May 18, 2026← Legal Center

Auto-apply features are not active in the current beta. This policy describes the safety framework that will govern automation when it is released.

Overview

JobCompanion is building toward controlled automation features that will allow users to delegate certain repetitive application tasks to AI-assisted workflows. We take automation safety seriously and have designed a multi-layer safety framework to ensure:

You always know what the system is doing on your behalf
Hard blocking rules prevent unsafe or unwanted actions
Every automated action is logged and auditable
You can pause, review, or cancel automation at any time
No automation proceeds without your explicit consent and configuration

Hard Blocking Rules

The following conditions will always cause automation to stop and alert you — regardless of your settings:

CAPTCHA Detection

If a CAPTCHA or bot-detection challenge is encountered during an automated workflow, the workflow immediately pauses. JobCompanion will never attempt to solve, bypass, or circumvent CAPTCHA systems. You will be notified and can choose to complete the application manually.

Login Wall Detection

If an application requires creating a new account or logging into a third-party platform, automation stops. We will not create accounts, enter credentials, or log into external platforms on your behalf.

Payment Page Detection

Any page requesting payment information, credit card details, or financial data immediately halts the workflow and flags it for your review. We will never enter payment information on any platform.

Unusual Form Detection

Forms with unusual fields (e.g., social security numbers, bank details, government ID numbers) trigger an immediate pause and human review requirement.

These blocking rules are hard-coded and cannot be disabled by users or administrators. They exist to protect you from fraud and unauthorized data submission.

Rule-Based Automation

JobCompanion's automation operates on a strict rule-based system, not open-ended AI decision-making:

Only jobs above your configured match score threshold are targeted
Only companies not on your blocked list are targeted
Only job types matching your configured preferences are targeted
Daily application limits cap the number of automated applications per day (default: 10)
Monthly limits apply to prevent runaway automation
Automation only fills fields that map to known, safe data from your profile
Unknown or ambiguous form fields are flagged for manual completion

Audit Logging

Every automated action is logged in your account's audit trail:

Timestamp of each automation event
Job listing targeted (title, company, URL)
Actions taken (form fields filled, documents submitted)
Blocking events (CAPTCHAs, login walls, payment pages encountered)
Approval or rejection decisions
Errors or failures with reason codes

Audit logs are retained for a minimum of 12 months and are accessible from your Activity page. You can export your audit log at any time.

Audit logs are stored securely and are only accessible to you and, in cases of abuse investigation, our safety team.

User Approval Systems

JobCompanion supports multiple levels of approval control:

Per-Job Approval Mode

Every job targeted by automation is presented to you for approval before any action is taken. You review the job, the generated application materials, and approve or reject each one individually.

Threshold-Based Approval

Jobs above a high confidence threshold (e.g., 90%+ match) can be configured to proceed automatically, while jobs below the threshold require manual approval. You set the threshold.

Batch Review

A daily digest of automation activity is available in your dashboard, allowing you to review what was done, approve pending items, and flag any issues.

Current Beta Status

In the current controlled beta:

Auto-apply is NOT active — no automated submissions occur
AI-assisted workflows require explicit review and approval at every step
Automation queue infrastructure is being built and tested
Safety rules are being validated before automation is enabled
Beta users can configure automation preferences but no automation runs

Current status: All applications require your explicit approval. Auto-apply is coming soon with full safety controls in place.

Future Safety Controls

Before auto-apply is released, we will implement:

Comprehensive testing across major job platforms (LinkedIn, Indeed, Greenhouse, Lever, Workday)
Independent safety audit of blocking rules and automation logic
User-facing safety dashboard showing automation health and recent activity
Anomaly detection to flag unusual automation patterns
Emergency kill switch to disable all automation platform-wide
Graduated rollout: per-job approval first, then threshold-based, then full auto
Clear user communication before each automation tier is unlocked

Questions about our safety approach? Contact us at our contact page.