No CV, no interview: what about feedback? What we know about OpenAI’s jobs platform
OpenAI’s AI jobs platform is back in the headlines as if it were the end of the CV and the interview, but it’s worth separating what’s confirmed from what isn’t. OpenAI announced it on 4 September 2025, and recent coverage, such as Panamericana‘s, presents it as a system that replaces CVs, interviews and applications with automatic matching. It matters because of a question almost nobody is asking: if an algorithm matches and rejects, what explanation does the person left out receive?
What has OpenAI actually announced, and what remains to be verified?
Here’s what can be verified: in its announcement, OpenAI talks about a Jobs Platform designed to connect workers with AI skills to companies that need them, along with certifications. The partners named include Walmart, John Deere, BCG, Accenture and Indeed, according to OpenAI’s own announcement. Outlets such as Digit.fyi reported that the launch was planned for mid-2026.
What we haven’t been able to confirm is the most eye-catching part of the headline. The idea of “no CVs or interviews” appears in later coverage, for example in Infobae, but in what we’ve reviewed of the original announcement the platform is described as a matching system, not as a promise to eliminate those steps. Nor have we found official confirmation of countries, dates or current availability. An analysis by Glozo, a company in the sector and therefore an interested party, argued that in August 2026 the platform still hadn’t launched, with no public beta or revised date. We haven’t found any launch announcement since.
What’s new in the story and what isn’t
The OpenAI Academy isn’t a recent development: according to The AI Insider, it already existed in 2025 as a free learning hub with more than two million participants, and the certifications are built on it. What was new was the commitment to certify 10 million Americans by 2030, which appears in the original announcement. In other words, the Academy as “training for those selected” is the press’s reading, not a literal description of the announcement.
What happens to the people who aren’t selected?
With or without a CV, any system that matches people also leaves people out, and that’s where almost all the public conversation falls short. People talk about how much time matching saves and whom it displaces, but say almost nothing about what the candidate who doesn’t fit receives. If the decision no longer goes through an interview, the natural moment when someone could say “this is what you were missing” disappears too.
A useful explanation for the candidate has three parts:
- Criterion: which skills or evidence were being sought for that role.
- Gap: where their application fell short or wasn’t visible enough.
- Next step: what they can do to improve, and how to ask for a review by a person.
A bare “you have not been selected” meets none of the three, whether it comes from a human or an algorithm.
What do EU rules say about explaining these decisions?
In Europe there’s a framework on the way, although it isn’t yet fully enforceable. The AI Act classifies systems for recruiting or selecting people as high-risk in its Annex III, including filtering applications and evaluating candidates, according to this analysis. The main obligations for that category apply from 2 December 2027, and from then on a company using one of these systems will have to inform the people affected.
Another legal analysis notes that the European Commission considers AI-based job profiling and matching tools will often be high-risk. The law doesn’t ban AI in hiring: it sets conditions for transparency and human oversight, as The Officer explains. We don’t know how OpenAI’s platform would fit into that framework, because there are no details on how it works or whether it will be available in the EU.
A caveat on the legal side
This summary isn’t legal advice. The Act’s application dates have changed with recent amendments, and its scope depends on each specific case. If you’re a candidate and believe an automated decision has affected you, or a company with questions about your obligations, consult an employment or data protection lawyer.
What can a candidate do today, whatever the platform?
Whether or not OpenAI’s platform arrives, many processes already use automated screening, and asking for an explanation is reasonable and simple. Three practical steps:
- Ask whether an automated system was involved. It’s a legitimate, neutral question, not an accusation.
- Ask for a specific, brief explanation. The easier it is to answer, the more likely you are to get an answer.
- Keep whatever you receive. Any criterion they give you is material for adjusting your profile for the next process.
You can use a sentence like this: “Thank you for the process. Could you tell me which skills or evidence would have strengthened my application, and whether any automated system was involved in the decision? It would really help me improve.”
In addition, asking for feedback after an interview is possible and simple. There’s no need to insist: a short, grateful message with one specific question is enough.
And if the system only matched, with no interview
If there was no interview, your best material is what the system could “see” of you. Check whether your profile shows results and context, not just job titles, and describe them with verifiable examples. If you say you can do something, make sure it can be checked.
What should hiring teams prepare?
Those who hire would do well to get ahead of this. An automated “no” with no explanation damages the candidate experience and the employer brand, even if the technology is good. To start:
- Identify at which stages of the process an AI tool is involved and what exactly it decides.
- Define in which cases a person reviews the result before it’s communicated.
- Prepare a three-line “no” message: criterion, evidence and next step.
- Tell the candidate clearly that an automated system is involved and how to request a review.
With this, there’s no need to wait for any obligation to give decent feedback: it’s good practice that also prepares you for what’s coming.
Where do we set the bar?
Our position is simple: getting rid of the CV may be progress if it widens who gets seen, but getting rid of the explanation for a “no” would be a step backwards, even with better technology. A matching platform should be judged by how it treats the rejected candidate, not just by how quickly it finds the chosen one. Until OpenAI confirms dates, scope and how it works, the prudent thing is not to take “goodbye to the interview” for granted, and to ask the same of any system, new or old: that whoever is left out knows why and what to do next.
Frequently asked questions
Has OpenAI launched its jobs platform yet?
We haven’t found confirmation of a launch. OpenAI announced it in September 2025 with mid-2026 as the target, and an industry analysis indicated that in August 2026 it still hadn’t launched. It’s worth checking OpenAI’s official information before taking anything for granted.
Will it really get rid of CVs and interviews?
That isn’t confirmed. The idea appears in several outlets’ coverage, but the original announcement describes a system for matching workers with AI skills and companies. There are no official details on how selection would work in practice.
What can I ask for if an automated system rejects me?
You can ask whether an automated system was involved and ask, briefly and kindly, which skills or evidence would have strengthened your application. If you believe the decision has harmed you unfairly, consult an employment or data protection lawyer.