Field notes, not marketing copy.
Reference pieces we add to when there is something worth saying, not a weekly feed. Written for whichever side of the desk you are sitting on.
Red flags when evaluating an AI ethicist candidate yourself
If you are vetting without us, here is what tends to separate a real practitioner from someone who has learned the vocabulary.
Automation bias: when a human “in the loop” stops actually deciding
A human reviewer sitting between a model and its consequence looks like accountability. It often is not, once the human has learned to defer to the system by default.
Placement, fractional, or rotation: a decision framework
The three shapes are not a menu to choose by preference. Here is the framework for matching the shape to the risk you actually carry.
Building evidence of competence without a certification to point to
EN 18274 has no sittable exam yet. Here is how practitioners build a credible evidence trail anyway.
Nine ways to regulate AI, sorted by whether they reach the person
UNESCO maps nine ways governments are choosing to regulate AI, from principles to liability. Read for compliance and they look similar. Read for who they actually protect, and the gap this business exists to close starts to show.
The standards behind the seat
Sentient Staff is a talent business, not a compliance desk built around one CEN spec. UNESCO, the OECD, the EU AI Act, EN 18274, and where each actually sits in the work.
AIGP vs EN 18274 vs nothing: what each actually signals
A sittable exam, a competence framework with no exam yet, and no credential at all, read for what a hiring manager should actually take from each.
The independence problem
Why AI ethicist roles fail quietly, not loudly, and the six things that decide whether a finding survives contact with a bad quarter.
Five things you will be told, and the clause that answers each
The predictable ways a finding gets waved off before it reaches anyone with the authority to act on it, and which of our six charter clauses exists specifically to stop each one.
The gap most AI ethics work never covers
Bias, privacy, and compliance are the easy part to check. What a system does to a person's attachment, trust, and sense of self is the part almost nobody evidences for.
Still deciding what you need?
Tell us what you are carrying and we will point you to the shape that actually fits, not just the one that is easiest to sell.