As AI takes on more decisions that affect people’s lives, from who gets a job interview to who gets a loan, one question keeps getting louder: who makes sure these systems are doing the right thing? Increasingly, the answer is an AI ethics specialist, a role that has moved from academic think tanks into compliance, product and risk teams.

The reason is practical rather than philosophical. AI systems that amplify bias, misuse personal data or make unexplainable decisions now create legal liability, regulatory scrutiny and reputational damage. Ethics has become an operational discipline with its own laws, standards and audit trails. This article explains what the role involves, why organizations need it, and how to cover it even without a dedicated hire.

Why the role emerged: harms that already happened

The case for AI ethics oversight is built on real incidents, not hypotheticals.

In 2018 Reuters reported that Amazon had scrapped an experimental recruiting tool after discovering it penalized résumés that included the word “women’s,” because it had learned from a decade of applications dominated by men. Nobody designed it to discriminate; it learned the pattern from historical data.

In 2023 the US Equal Employment Opportunity Commission settled its first case involving AI-driven hiring discrimination. The online tutoring company iTutorGroup agreed to pay $365,000 after the EEOC alleged its application software automatically rejected older applicants.

These examples share a lesson. Bias in AI usually comes from data and design choices that nobody questioned, and it scales instantly. A biased human reviewer affects dozens of decisions; a biased model affects thousands before anyone notices.

What an AI ethics specialist actually does

AI ethics specialists sit at the intersection of technology, law and the people affected by automated decisions. Their mission is to make sure innovation and integrity move at the same speed. In practice, the work falls into three areas.

Building governance frameworks. They define how AI is proposed, reviewed, tested, deployed and monitored inside the organization. That includes a register of AI systems, risk tiers that decide how much review each use case needs, requirements for human oversight, and data privacy standards.

Testing for bias and harm. Working with engineers and data scientists, they examine training data for gaps and skews, test model outcomes across demographic groups, and look for harms beyond bias, such as privacy leakage, unsafe outputs or manipulation. The goal is AI that performs fairly, not just accurately.

Translating between code and accountability. They explain how a model reaches decisions, in terms that executives, regulators, customers and affected individuals can understand, and document it so the organization can defend its choices later.

The laws and frameworks behind the role

A growing body of law and standards gives the work structure and teeth:

• New York City Local Law 144, enforced since July 2023, requires employers using automated employment decision tools to commission annual independent bias audits and notify candidates.
• The EU AI Act, in force since August 2024, classifies AI used in hiring, credit, education and essential services as high-risk, with obligations for risk management, data governance, documentation and human oversight phasing in over several years.
• Colorado’s Artificial Intelligence Act, passed in 2024, was the first broad US state law aimed at preventing algorithmic discrimination in high-risk AI systems.
• The NIST AI Risk Management Framework offers a voluntary structure for mapping, measuring and managing AI risks.
• ISO/IEC 42001, published in 2023, is the first international management system standard for AI, allowing organizations to certify their AI governance.

Existing laws still apply too. Anti-discrimination, consumer protection and privacy laws do not have exceptions for decisions made by software.

How to build AI ethics into your organization

Not every organization needs a full-time AI ethics specialist, but every organization using AI in decisions about people needs the function. A practical starting sequence:

1. Inventory your AI use. List every system that uses AI, including features embedded in vendor software such as applicant tracking, credit scoring or customer service tools.
2. Tier by impact. Flag the systems that make or influence decisions about employment, credit, housing, health, insurance or access to services.
3. Ask vendors hard questions. Request bias audit results, documentation of training data and details of how decisions can be explained and appealed.
4. Test outcomes, not intentions. Compare results across groups, and repeat after significant model or data changes.
5. Keep a human in the loop for high-impact decisions, with the authority and time to override the system.
6. Document and review. Record decisions and reasoning so you can show regulators, auditors or courts what you did and why.

Ethics as an enabler, not a brake

A common worry is that ethics review slows innovation. In practice, the opposite is usually true. Projects that address bias, privacy and explainability early avoid late-stage surprises that kill launches, and they reach regulated markets faster because the documentation already exists. Customers and partners increasingly ask about AI governance in procurement questionnaires, and organizations with good answers win those deals.

The companies that embed ethics into their AI strategies now will be the ones customers, employees and regulators trust later. The question every leader using AI should be able to answer is simple: who is making sure our technology aligns with our values and our legal obligations?

Frequently asked questions

What background does an AI ethics specialist need?

The role draws people from law, compliance, data science, philosophy, social science and product management. What matters most is the ability to understand how AI systems work technically and to connect that to legal requirements and human impact.

Do small businesses need to worry about AI ethics?

Yes, if they use AI in decisions about people. Many small businesses use AI through vendor tools for hiring, lending or customer service, and they remain responsible for discriminatory outcomes even when the vendor built the model.

Is AI ethics the same as AI compliance?

They overlap but are not identical. Compliance focuses on meeting legal requirements. Ethics also asks whether something should be done even where the law is silent, which matters because AI law is still catching up with the technology.

Responsible AI, built in from the start

Delana Technologies helps organizations inventory their AI use, assess risk, and put governance and compliance controls in place that stand up to scrutiny. Learn about our cybersecurity and compliance services and AI consulting, or read Shadow AI in 2026 for the governance gaps most organizations miss. To start the conversation, call 239.414.5126 or contact us.

---

Sources: Reuters reporting on Amazon’s experimental recruiting tool (October 2018); US EEOC announcement of the iTutorGroup settlement (2023); NYC Department of Consumer and Worker Protection guidance on Local Law 144; EU AI Act (Regulation (EU) 2024/1689); Colorado SB 24-205; NIST AI Risk Management Framework (AI RMF 1.0); ISO/IEC 42001:2023.