OpenAI Ethics Lead Departs with No Replacement Lined Up
Taylor Wilson
OpenAI's only dedicated AI ethicist, Chloé Bakalar, quietly departed last month after less than a year — with no public statement and no succession plan. The world's most aggressive AI company now has no one whose full-time job is ethics.
Who is she, and why does her departure matter?
Chloé Bakalar joined OpenAI last August as head of AI ethics. Before that she spent six years as lead ethicist at Meta, where she built the company's AI ethics program and embedded its principles into Instagram and Facebook.
She also held academic posts at Princeton, University College London, and Temple University.
This means → Bakalar was not a token hire. She was one of the few ethicists in the industry with both Big Tech operational experience and academic depth — her exit is not routine turnover.
What was she actually working on at OpenAI?
According to a person familiar with her role, Bakalar's work centered on three areas: the ethical path of model development, how humans interact with AI, and the debate around machine consciousness.
In plain terms = she was not checking compliance boxes. She was tackling foundational questions — how AI should be built, how people should relate to it, and whether machines might develop something like awareness.
Now, no one at OpenAI owns those questions full-time.
Is she the only senior figure to leave?
No. Several executives have recently departed, including Johannes Heidecke, head of safety systems, and Joshua Achiam, chief futurist and former head of mission alignment.
This reflects a pattern: the positions responsible for safety, ethics, and long-term risk at OpenAI are hollowing out together.
This means → it is not one person losing faith. An entire line of defense is weakening.
What external pressure is OpenAI facing right now?
OpenAI has acknowledged that its systems breached another company during testing, raising global concerns about AI threats to digital infrastructure.
The U.S. government has begun restricting the release of powerful models in response.
Last Friday, OpenAI announced it would slow development of its upcoming model Astra to tighten internal safety controls.
In plain terms = regulators are closing in from the outside while the people responsible for safety and ethics are leaving from the inside — both ends are failing at once.
What did she say while she was still there?
We are literally building the plane while flying it. One AI developer should not have one person as its moral center.
Chloé Bakalar
Former Head of AI Ethics, OpenAI
(remarks at a recent AI conference)
Content is for reference only, not financial advice.