AI & Technology Intelligence
AI NEWS INSIDER
Issue #65 · June 29, 2026
THIS WEEK: An AI agent ran 70,884 real job interviews, and the results map out exactly where AI delivers. Here is what that tells you about which work to hand to AI, and how to put it to use in your own role first.
6 MIN READ | Sharp, actionable intelligence
What 70,884 AI Job Interviews Reveal About Where AI Actually Works
By AI News Insider Editorial · 5 min read
A study making the rounds this month should stop every manager cold. An AI voice agent named Anna conducted 70,884 real job interviews across 48 openings inside a live hiring pipeline. And the results were strong: the AI-led process drove double-digit gains in offers accepted, candidates who actually started, and retention months later. The economists behind it, Brian Jabarian of Chicago Booth and Luca Henkel of Erasmus, ran it as a genuine field experiment, not a lab demo.
The number that surprised even the researchers
It is not the offer rate. It is this: when candidates were free to choose how they interviewed, 80% opted into the AI. They were not pushed into it. They chose it, and they cited three reasons: it felt fairer, more consistent, and more convenient. The takeaway is not about who does the job better. It is that people are far more ready to work alongside AI than most leaders assume, and that readiness is the real story of 2026.
Why it worked, and what made the difference
This is the part to internalize. The result did not come from raw intelligence. It came from three things the AI did well at scale: it asked every candidate the same questions the same way, it never had an off day across 70,000 conversations, and it was available whenever the candidate was. Candidates valued exactly those qualities. "The attractiveness of AI is that it's always available," Jabarian put it. Consistency was the advantage. That is the pattern worth remembering.
Why this lands on your desk, not just HR's
Forget recruiting for a second. The study is really a map of where AI delivers first, in any function. It is the high-volume, standardized work where a consistent process matters and being available around the clock is itself an advantage. That describes pieces of sales, support, ops, and screening in almost every company. If part of your role is repeatable, scriptable, and judged on consistency, that is exactly the slice AI is ready to take on. The move is not to fear it. It is to read the map and decide what to hand over, so you can spend your hours on the work that consistency cannot replicate.
The Bottom Line for AI News Insider Readers
70,884 interviews make one thing clear: AI delivers most where work is standardized, high-volume, and judged on consistency. That is the tell for where it pays off next, whether you are deciding what to automate or where to focus your own energy. The S.C.A.N. test below turns that one insight into a checklist you can run on any task in your week.
The recruiter study is not really about recruiting. It is a test for spotting which work AI is ready to take on, before everyone else figures it out. Four conditions made this deployment work. When a task has all four, it is a strong candidate for AI. SCAN any task in your week against them.
S. Standardized
The AI asked every candidate the same questions, scored against the same criteria, every time, with no drift. Any task with a repeatable script and clear criteria is a strong fit. If a task's value is in doing it the same way reliably, that is exactly where AI is strongest. The work that stays yours is the opposite: where the right answer changes with context.
C. Convenient at scale
Anna ran 70,884 interviews without tiring, at any hour the candidate wanted. Volume and round-the-clock availability are where AI shines. When a task is high-volume and "always on" matters more than a personal touch, it is a natural fit for AI. Ask of any task: does it get better with a personal relationship, or just faster with availability?
A. Adjudication stays with a person
Read the study carefully and the people did not vanish. The AI ran the interview; a person still owned the final hiring decision. That is the durable shape: AI handles the repeatable interaction, a person owns the consequential call. The roles that stay valuable are the ones that move up to the judgment, not the ones that cling to the script.
N. Neutrality builds the trust
Part of why candidates embraced the AI was that it felt fair and consistent, the same process for everyone. Where people value an even-handed, standardized experience, an AI process can build trust rather than erode it. If a task is one where consistency and fairness matter most, that is a strong signal it is a fit for AI.
Your Monday Morning Action
Block 20 minutes and SCAN your own role. List every recurring task you did last week. Tag each on the four conditions: Standardized, Convenient at scale, Adjudication (does a person own the call), Neutrality. Anything that scores high on Standardized plus Convenient is a strong candidate to hand to AI, so either automate it yourself this quarter and claim the win, or move your hours toward the Adjudication work that needs you. The opportunity goes to whoever reads the map first.
Hiring
The AI Recruiter Stopped Being a Pilot. In 2026 It Became a Hiring Trend.
Voice-AI screening is spreading fast across high-volume hiring, while external hiring slows and internal mobility climbs about 6% as companies redeploy talent instead of backfilling. If you are job hunting, expect a machine on the first call. If you are hiring, the question is no longer whether to use AI for screening, but where a human still has to own the decision.
Big Tech
Salesforce Cut Support From 9,000 to 5,000 Roles. The AI Excuse Is Now on Earnings Calls.
Marc Benioff said Salesforce shrank internal support from 9,000 to 5,000 roles, with AI resolving 85% of service inquiries, and will not backfill those seats. TechCrunch's running tally now counts 75,000+ layoffs in 2026 where employers cited AI, including Oracle's 21,000 and Meta's 8,000. The "no longer backfilling" line is the one to watch. It signals categories closing, not just slower hiring.
Enterprise
Spend on AI Agents Hits $206.5B. Gartner Says 40% of Projects Still Get Killed.
AI agent software spending is projected at $206.5 billion in 2026, up about 139% in a year, yet Gartner expects more than 40% of agentic projects to be canceled by 2027 on unclear value and cost. The winners look like Anna: one clear job, a measurable result. The casualties are vague "do everything" agents nobody can put a number on.
Infrastructure
ServiceNow and NVIDIA Launch Project Arc. The Cure for Agent Chaos Is Governance.
At Knowledge 2026, Jensen Huang and Bill McDermott unveiled Project Arc, a long-running desktop agent built on NVIDIA's OpenShell, an open runtime that sandboxes what an agent can see and do. The pitch is not a smarter agent. It is a governed one, with a human still in control of the consequential decisions.
70,884
Job Interviews the AI Agent "Anna" Conducted
Across 48 openings inside a live hiring pipeline. A real field experiment at full scale, not a demo.
80%
Of Candidates Opted Into the AI Interview
When free to choose how they interviewed. They cited fairness, consistency, and convenience. They were not pushed into it. They chose it.
+18%
Lift in Candidates Who Started the Job
After the AI-led process, an 18% gain in the study. The same process also lifted accepted offers by 12%.
+16-18%
Higher Retention Months After Hire
The AI did not just fill seats faster. The people hired through it stuck around longer, up to four months out. Consistency drove the result.
Metaview (AI for Structured, Fairer Hiring)
Bring structure and consistency to every interview your team runs
What it is: An AI recruiting assistant that listens to your interviews, writes the notes for you, and turns each conversation into a structured candidate evaluation scored against your rubric, not a generic summary. It plugs into the calls and tools you already use and hands every interviewer the same scaffolding.
Why it matters now: This is the "S" and the "N" of the S.C.A.N. test, applied to a process you still run yourself. The study showed how much standardization and a consistent, fair experience matter. Metaview pushes your interviews toward both: same questions, same scoring, far less interview-to-interview drift. You keep a person in the chair and get the consistency that makes a process work.
Who gets it: Founders and hiring managers who run interviews themselves and want them to be fairer and more comparable without buying enterprise recruiting software. It is built for teams of any size, with a free way to try it before you commit. Start by standardizing your scorecard on one open role, then let the tool hold every interviewer to it.
"The attractiveness of AI is that it's always available."
Brian Jabarian, Chicago Booth · co-author of the 70,884-interview field experiment, on why candidates embraced the AI interview · 2026
Share AI News Insider
Know someone hiring or job hunting this year? Forward them this issue.
SHARE & EARN REWARDS →AI News Insider · Your weekly edge in artificial intelligence
aiinsider247.com