Five values replace a longer collection of principles
OpenAI has released a substantial revision of its public Model Spec, the document that describes how its models are intended to behave. The new structure centres five values: faithfulness to human intent, harm minimisation, honesty and transparency, human agency and care. OpenAI says the change is meant to make difficult trade-offs easier to understand as systems become more capable.
The values are not a simple ranking that resolves every conflict. A useful assistant may need to refuse a harmful request while still explaining the boundary honestly and preserving the user’s ability to choose a safer course. Product teams therefore need concrete evaluations for situations where helpfulness, privacy, safety and institutional instructions pull in different directions.
Agentic work makes context and oversight more important
The update gives more attention to context, oversight and agency. An agent working across files, tools or long-running tasks needs to understand who authorised the work, which instructions remain current and when a human expects to review the result. That is materially different from answering a single question because a mistaken assumption can propagate through several actions before anyone notices.
OpenAI also changed several behavioural defaults and retired much of the detailed style guidance previously held in the specification. The shift recognises that tone and formatting are often product or user choices, while the more durable questions concern truthfulness, authority and consequence. Developers should still test that model updates do not unexpectedly change the voice or conventions their users depend on.
A public specification is a testable commitment
Publishing the specification creates a reference for researchers, customers and affected users, but it does not prove that every deployed model follows it perfectly. The practical test is whether behaviour can be measured against the stated values across languages, adversarial prompts and real workflows. Material gaps should be documented, prioritised and retested after model or system changes.
For organisations building agents, the revision is a prompt to translate broad values into operational controls. Tool permissions, confirmation points, evidence citations, escalation rules and audit records should express what human agency and honest uncertainty mean in the actual service. A value becomes useful when it changes a system decision and leaves evidence that the decision can be reviewed.
What teams should validate before production
For implementation, teams should record the exact model, software, dataset, connector and policy versions behind every result. Representative evaluation must include normal work, boundary cases, deliberate misuse and component failure, with outcomes separated by user role, data sensitivity and consequence. A strong average should never hide a serious failure in a critical path. Changes should pass the same regression set before promotion, and operators need a documented fallback that preserves evidence rather than silently broadening access or lowering quality.
Procurement and planning should account for the human work required to review exceptions, maintain controls and investigate incidents. Published performance or efficiency claims are useful starting points, but production confidence comes from evidence collected in the organisation’s own environment. A named owner, measurable acceptance threshold, cost model and tested shutdown path turn a promising release into a service that can be governed over time. Where decisions affect safety, rights, money or regulated records, accountable professional review remains essential.
Security and operational evidence
Security review should map every identity, credential, network destination, stored artefact and administrative access path involved in the workflow. Logs need enough context to reconstruct who initiated an action, which evidence was consulted, what the system proposed and what ultimately executed. Sensitive prompts and traces should not be retained merely because they are useful for debugging. Retention, redaction and access must follow the underlying data classification, with a practical process for correcting or removing material when policy or law requires it.
Operational monitoring should measure quality and service behaviour together. Latency, throughput and availability matter, but so do unsupported claims, missed constraints, unsafe tool calls and the rate at which people overturn recommendations. Teams should rehearse dependency outages and partial failures before unattended operation. If a model, connector or external service becomes unavailable, the workflow should fail predictably, preserve its state and tell the user what remains unfinished. Clear degradation is safer than an agent improvising around a missing control.
Transparency should scale with adoption
Organisations should explain the scope and limitations of the deployed capability to the people affected by it. Users need to know when automation is involved, how to challenge an outcome and where a human can intervene. That transparency improves feedback and prevents a pilot result from being interpreted as blanket approval for broader autonomy. Expansion should follow demonstrated evidence, not novelty or pressure to match a competitor announcement. The durable advantage comes from dependable integration, careful measurement and learning that survives the next model or product update.
Finally, teams should schedule a review after real usage rather than treating launch as the end of governance. The review should compare expected benefits with observed quality, cost, incidents and user behaviour, then retire workflows that do not justify their complexity. A service that works well can expand through the same evidence-based process. A service that fails should be narrowed or stopped without losing the records needed to understand why. This feedback loop is what turns experimentation into accountable operational capability.