Position papers, incident analysis, framework assessments, and monthly intelligence that make production AI safety inspectable. The work is public because a standard only matters if people can inspect it.
Five developments shaping production AI this month — with the PAI angle on what they mean for practitioners building real systems today.
Read Issue 001 →Get the next issue in your inbox
Free. Monthly. Unsubscribe anytime.
Research creates leverage only when the next step is obvious. Pick the route that matches the pressure you are under.
Executive pressure
Use this when board, procurement, or governance teams need a credible path from research into rollout.
Choose org path →Live system scrutiny
Use DSA when an AI workflow is already making decisions, touching customer data, or creating operational risk.
Scope a DSA →MSP client demand
Turn the research library into client discovery, policy workshops, and monitored controls your team can deliver.
Get MSP launch pack →Individual credibility
Start with free AIDA, then move into specialist certification when your role demands stronger evidence.
Start free AIDA →Structured reliability testing of frontier AI models and agent frameworks against PSF criteria. Quarterly scorecards. Open methodology.
Explore →
Analysis, essays, and deep-dives on AI deployment, safety, and the production practitioner experience.
Explore →
Reusable workflow patterns for production AI systems — vetted against the PSF and ready to adapt.
Explore →
Documented AI failure cases with root-cause analysis mapped to PSF domains. Learn from what went wrong.
Explore →
Independent PSF assessments of every major AI framework — LangChain, CrewAI, AutoGen, Cursor, and more.
Explore →
The Production Safety Framework itself — eight domains, openly published and freely referenceable.
Explore →
Generate a PSF-aligned readiness report for an AI agent, with evidence grade, repository signals, and a shareable badge.
Explore →
May–June 2026 lab and ecosystem work — indexed here for procurement visitors; formal publication titles below are unchanged.
Multi-team governance, org-level IdP, and usage analytics for Cursor Enterprise Organizations GA (3 Jun 2026).
Read →Enterprise plugins, hosted Sites, and human-refinement signals for Codex knowledge work (2 Jun 2026).
Read →AWS-native governance and regional inference for OpenAI models and Codex on Bedrock GA (1 Jun 2026).
Read →PSF Workflow Studio released under MIT — the working artifact behind the open Production Safety Framework text.
Read →Empirical scan of 20 public agent repos against PSF evidence signals (PAI-ARI-2026.1).
Read →Vendor-reported ChatGPT, login, and checkout failures on 29 May 2026 — mapped to D8, D4, and D5; indexed in the incident registry.
Read →Verified May 2026 filing failure mapped to D2, D5, and D6 — indexed in the incident registry.
Read →Weekly practitioner index of vendor and product data-use posture changes.
Read →Independent PSF coverage review of Codex CLI for production agent workflows.
Read →Vendor resilience and deployment-safety signals for Cursor Automations 3.5.
Read →Framework assessment of Google Agent Executor against PSF domains.
Read →Position papers, analyses, and framework notes from the PAI research programme.
Maps PSF domains to EU AI Act obligations for high-risk AI system deployers. Covers conformity assessment requirements, technical documentation standards, and human oversight obligations.
Analysis of common failure modes in production LLM deployments. Identifies root causes across PSF domains and intervention patterns.
Examines what constitutes meaningful human oversight in high-stakes AI-assisted decisions. Includes design patterns for effective human checkpoints.
Documents the reasoning behind each PSF domain, alternatives considered, and how practitioner feedback shaped the framework.
Patterns repeatedly observed in anonymised production assurance reviews and incident-led postmortems:
Model text consumed by downstream systems as if it were trusted structured data.
Review PSF-D2 →Strong model-call logs, weak cross-service traceability at queue and handoff boundaries.
See Lab methodology →Operational actions executed without explicit human intervention criteria on high-consequence paths.
Use deployment guidance →PAI collects anonymised incident reports from practitioners to inform framework development. If you have experienced a production AI failure and are willing to share details, we welcome your contribution.
PSF updates, deployment checks, failure patterns, and proof paths for practitioners, MSPs, and teams who need AI work to survive scrutiny. No hype.