AI hiring tools: what candidates should know

The main types of AI hiring tools employers use

Employer-side AI falls into a few distinct categories. Knowing the difference matters because each tool touches your application at a different point and can be challenged or worked around in different ways.

Applicant tracking systems (ATS) have existed since the 1990s, but modern versions now include AI ranking layers that score resumes before a human ever opens them. Video interview platforms use computer vision and natural language processing to score your facial expressions, word choice and speaking pace. Chatbot pre-screeners ask you a set of questions and filter you out if your answers fall below a threshold. Predictive hiring models pull in data from your application, public profiles and sometimes third-party data brokers to estimate your likelihood of success or retention.

• ATS with AI ranking: used by most companies with more than 50 employees; scores resume keyword density, formatting and job-title proximity
• Video interview AI: platforms such as HireVue and Spark Hire analyse recorded answers; some score tone and micro-expressions
• Chatbot pre-screeners: common on high-volume roles; ask knockout questions about salary, location and eligibility before routing you forward
• Predictive models: aggregate signals across your profile to produce a fit score; inputs are often opaque

How ATS ranking actually works

Most ATS platforms parse your resume into structured fields: job titles, employers, dates, skills and education. The AI layer then compares those fields against the job description using a combination of keyword matching and semantic similarity. A resume that uses the exact phrase from the posting will generally score higher than one that uses a synonym, even if the meaning is identical.

Formatting matters as much as content. Multi-column layouts, text inside graphics, headers in text boxes and unusual fonts can all cause parsing errors. When the parser fails, fields are left blank and your score drops. A clean, single-column resume in a standard font is not a stylistic preference; it is a technical requirement for getting past the first filter.

The job description itself is your best guide. Read it carefully and mirror the language it uses for skills and responsibilities. If the posting says 'stakeholder management', use that phrase rather than 'managing stakeholders'. This is not keyword stuffing; it is speaking the same language as the system scoring you.

Video and voice AI: what is being scored

Several large employers use asynchronous video interviews as a second filter. You record answers to preset questions and the platform scores them before a recruiter watches. The scoring criteria vary by vendor, but commonly include speaking pace, vocabulary range, filler word frequency and, on some platforms, facial movement analysis.

You have practical options here. Speak at a measured pace, around 130 to 150 words per minute. Use specific examples rather than general statements. Look directly at the camera, not at your own image on screen. Avoid long pauses at the start of answers; the system may interpret silence as low confidence.

In the UK, the Information Commissioner's Office has stated that automated video scoring that produces a decision without human review may require explicit consent under UK GDPR. In the US, Illinois passed the Artificial Intelligence Video Interview Act in 2020, requiring employers to disclose AI use, explain how it works and get consent before using video AI on Illinois residents. No equivalent federal law exists yet, but several states are considering similar legislation in 2026.

Your legal rights by country

The legal landscape is uneven but moving quickly. Knowing your rights in your jurisdiction lets you ask the right questions and, where necessary, push back.

• United States: no federal AI hiring law as of October 2026; Illinois AI Video Interview Act covers video scoring; New York City Local Law 144 requires bias audits for AI tools used in hiring decisions affecting NYC residents; Colorado and California have pending or recently passed algorithmic accountability bills
• Canada: the federal Privacy Commissioner has said that automated hiring decisions may require meaningful human review under PIPEDA; Quebec's Law 25 (in force since 2023) gives individuals the right to know when a decision is made solely by automated means and to request human review
• United Kingdom: UK GDPR Article 22 gives you the right not to be subject to a decision based solely on automated processing if it produces a legal or similarly significant effect; you can request human review and ask for an explanation of the logic involved
• Australia: the Privacy Act 1988 does not yet include a specific automated decision right, but the government's 2023 review recommended adding one; employers using AI scoring are expected to disclose this in their privacy notices under current guidance

Questions you can ask employers about their AI tools

You are entitled to ask about the tools being used on your application. Most recruiters will answer straightforward questions, and in some jurisdictions they are legally required to. Asking also signals that you are informed, which can work in your favour.

Send these questions by email so you have a written record. Keep the tone neutral and professional.

1. Does your process include any automated screening or scoring of resumes before human review?
2. Is there an AI component to the video interview stage, and if so, what factors does it score?
3. Can I request a human review of any automated decision that affects my application?
4. What third-party data, if any, is used alongside my application materials?

How to protect yourself when applying

The most effective defence against opaque AI screening is a resume that is both machine-readable and genuinely accurate. Tailoring your resume to each job description is no longer optional; it is the baseline for getting past automated filters.

Break-In Bad reads the full job posting, including location, visa requirements, time zone and seniority level, before deciding whether to apply on your behalf. For roles scored between 45 and 74 percent, it flags them for your review rather than auto-applying, so you are not wasting applications on poor fits. When it builds or enhances your resume, it checks every line against what you have confirmed and keeps your original, so nothing is invented. That matters when an employer's AI is cross-referencing your resume against LinkedIn or a background check.

Beyond tooling, keep a record of every application you submit, including the job description as it appeared at the time. Job postings are sometimes edited after you apply, and having the original text protects you if a discrepancy is raised later.

Red flags in job postings that suggest heavy AI filtering

Some postings are written primarily to feed an ATS rather than to attract candidates. Recognising them helps you decide how much effort to invest and how to tailor your materials.

A posting that lists 20 or more required skills, uses identical phrasing across multiple job titles at the same company, or asks you to paste your resume into a text field rather than upload a file is likely running heavy automated screening. In these cases, matching the exact language of the posting is especially important, and a cover letter may carry less weight than usual because it may not be parsed at all.

• Very long skill lists with no distinction between required and preferred
• Generic language that could apply to any company in the sector
• Application portals that strip formatting from uploaded documents
• Immediate automated rejection emails, which suggest knockout question filtering rather than resume review