AI Interviewers Are on Sale. Read This Before One Screens Your Next Hire.
Tools that rank resumes and hold the first interview are reaching small hiring teams. The law treats their choices as yours, so set the rules first.
By John Martines · October 14, 2026 · 6 min
HackerRank just put its AI interviewer on general sale, and the job sites you already use offer AI ranking. Here is what the hiring laws ask of a small employer, and five questions to ask before any tool screens a candidate.
The owner of a 20-person heating and cooling company needs two service technicians before winter. The job post went up Monday. By Thursday there are 140 applications, and the job site offers a helpful button: let AI sort them and show you the best matches first. Another vendor emails the same day. Its AI will hold the first interview for you, by video, and send a score.
Both would save her a weekend. Both also make a decision about real people that she will have to answer for, and most owners don't know that yet.
AI hiring tools come in a few shapes. Some read resumes and rank them. Some chat with applicants and ask screening questions. Some record a one-way video interview and score it. The newest ones hold the whole conversation: they ask questions, follow up on the answers, and hand you a scorecard.
On October 5, 2026, HackerRank made its AI interviewer, called Chakra, generally available after a six-month test that ran about 500,000 interviews, with Snowflake and Capgemini among the early users. It is built for hiring software engineers, and HackerRank says the final hiring decision stays with people. It won't be pitched to a heating contractor. But it shows where the tools are heading, and the simpler versions are already inside the job sites and hiring software small businesses use.
The legal idea to hold on to is simple. If a tool screens out a candidate for you, the law in more and more places treats that as your decision. Not the vendor's.
How it got here:
- January 1, 2020: Illinois's AI Video Interview Act took effect. Employers using AI to judge video interviews must tell applicants, explain what it looks at, and get their consent first.
- July 5, 2023: New York City began enforcing Local Law 144. A tool that scores or screens candidates needs a bias audit within the past year, a public summary of the results, and notice to candidates at least 10 business days before it is used.
- May 2025: A federal judge let a lawsuit against Workday go forward as a nationwide collective action on behalf of applicants 40 and older, who claim its AI screening rejected them. The court had earlier accepted that a screening vendor can act as the employer's "agent."
- October 1, 2025: California rules took effect for employers with five or more workers who use automated tools in hiring and other job decisions: keep records for four years, and the vendor screening for you counts as your agent.
- January 1, 2026: An Illinois law took effect for employers with even one Illinois employee: AI in job decisions may not discriminate, ZIP codes may not stand in for protected traits, and workers and applicants must be told when AI is used.
- October 5, 2026: HackerRank's AI interviewer went on general sale.
Big companies have HR departments and lawyers to vet these tools. A small business usually has the owner and an office manager, and the tool shows up as a checkbox.
The blame doesn't move to the vendor. California's rules, New Jersey's rules and the Workday case all point the same way: when a vendor screens for you, you share the responsibility for who got turned away.
Size doesn't exempt you. Illinois covers employers with one employee. California's rules start at five. New York City covers anyone hiring for a job there.
Good people disappear quietly. A rejected candidate never tells you the tool got it wrong. They just apply somewhere else. A technician with ten years of experience and an unusual resume can be ranked under people who used the right words.
AI companies have rules about this too. Anthropic's usage policy, in a new version that starts November 12, 2026, lists "screening and hiring" as a high-risk use: a qualified person must review the recommendation and the candidate must be told AI was used. Read more in [Claude's New Rules Start November 12](/articles/claude-rules-november-12).
None of this means don't use AI in hiring. It means use it for the parts that are safe, and keep a person on the parts that decide.
A free account opens the rest of this article and the whole library.