AI is transforming every industry. Customers, employees and other stakeholders increasingly value organisations that use it responsibly while remaining accessible, accountable and committed to investing in people.
We’re not anti-AI. Many organisations use AI to work more efficiently and serve people better. But adoption has often moved faster than transparency and accountability. People cannot always tell when they are interacting with AI, how it influences decisions, or whether meaningful human support is available when something goes wrong.
PeopleFirstVerified is being developed as an independent verification standard. It will give organisations a credible way to demonstrate that their use of AI is built on transparency, meaningful human access and clear accountability. It will also give customers, employees, investors and other stakeholders a straightforward signal of that commitment.
What we verify
Transparency
People should be able to understand when and how AI is being used, particularly when it interacts with them or influences important decisions.
- Clear disclosure of relevant customer-facing and decision-making uses of AI
- Plain-language explanations of what AI does—and does not—decide
- Appropriate disclosures under Article 50 of the EU AI Act, where those requirements apply
- Clear information about significant limitations or circumstances in which human involvement may be required
Human access
People should have a clear and meaningful route to human assistance, especially when they need support, wish to make a complaint, or are affected by an important decision.
- An accessible route to a real person
- A defined process for escalating complaints or significant concerns
- Clear information about expected response times
- Appropriate opportunities to request human consideration of AI-influenced decisions
Accountability & governance
Responsibility for an organisation’s use of AI should remain with identifiable people—not with the technology itself.
- Clear internal ownership of AI-related decisions and outcomes
- A process for questioning or challenging significant AI-influenced decisions
- Proportionate review of AI systems for performance, errors, fairness and unintended effects
- Appropriate oversight based on the nature and potential impact of the AI involved
How the assessment works
PeopleFirstVerified uses artificial intelligence to review publicly accessible website pages and linked documents across transparency, human access and accountability. The resulting report is a dated assessment of what an external visitor could reasonably identify at that time. It does not examine internal systems, test whether published commitments operate in practice, or determine legal compliance. Absence of public evidence does not necessarily mean that a policy or safeguard does not exist.
Where public evidence indicates that Article 50 of the EU AI Act may apply, the assessment considers whether relevant disclosures are visible. It does not determine whether the organisation is legally subject to Article 50 or compliant with it.
We’re developing the standard with input from the businesses, customers, employees and investors who may rely on it.
You can run a preliminary automated assessment of a public website here. It provides an AI-produced overview based only on publicly accessible website content and documents. It is not formal PeopleFirstVerified verification, certification or legal advice. Nor has it been reviewed by a human at this stage.
Register your interest to help shape the standard before launch.
No cost. No obligation. We will only contact you about PeopleFirstVerified updates