Leadly

AI Policy

Last updated July 2026

Leadly uses AI in a small number of specific features. This page lists every one of them, what data each sends to our AI provider (Anthropic), and the limits we place on automated decisions. If it isn't listed here, it doesn't use AI.

Job post safety screening

When a job is posted, it's screened automatically. Clear-cut violations (drugs, sexual content, weapons, violence, advance-fee scams) are blocked by fixed keyword rules — not by AI. Common, obviously-safe local jobs (babysitting, dog walking, lawn care, tutoring and similar) publish immediately without AI review.

For everything else, an AI classifier reads the job's title, description, and instructions. It is deliberately limited: it can mark a post 'safe' or send it to human review — it can never reject a post on its own. If the AI is unavailable, screening falls back to keyword rules alone. Screening decisions are logged for accountability.

Find Clients (freelancers)

The opportunity engine sends public business information (from Google Places — name, category, rating, website) together with your stated skills, age range, experience, and city to the AI, which scores how good a fit each business might be and drafts outreach (emails, DMs, call scripts, project plans). Drafts use placeholders for your name and are filled in on your device.

AI opportunity scores and drafts are suggestions, not facts. They can be wrong or generic. Review and edit everything before sending — you are the sender, not the AI.

Application helper

If you ask for help applying to a catalog job, the AI is sent the job name and your age range, and returns tips: sample resume bullets, likely interview questions, what to wear, and how to apply. Nothing from this helper is stored — it's generated for you on the spot.

What Leadly's AI does NOT do

There is no AI matching or ranking of job seekers: the Find Work feed is ranked by distance, and search is keyword-based. AI does not read your private messages (message scam warnings use fixed keyword patterns, on your device's session — not AI). AI does not make account decisions: suspensions and bans are human actions, logged with the acting moderator.

What data goes to the AI provider

Only what each feature needs, as described above: job-post text for screening; skills/age range/experience/city plus public business data for Find Clients; job name plus age range for the application helper. We never send your email, phone number, password, private messages, or exact location to the AI provider. Processing happens under our provider's commercial API terms.

Content we create with AI

Some site content (for example certain blog or landing pages) may be drafted with AI assistance and reviewed before publication. Blog guides are general information, not professional advice.

Questions about AI features: [email protected].