April 2026: bulk access reviews, FAIR risk scoring, and integration error UX¶
Thalian now supports bulk decisions in access review campaigns, a FAIR-aligned entity risk model, and smarter integration error handling.
Bulk access review decisions¶
Access review campaigns now support bulk approve and bulk revoke. Select multiple items, apply a decision, and move through large reviews faster. Campaigns past their due date automatically send overdue reminder emails to reviewers with a direct link to the campaign.
FAIR-aligned entity risk scoring¶
Entity risk scores have been redesigned around a FAIR-aligned model that weighs privilege level, platform exposure, finding severity, and behavioral signals. Scores are more meaningful and comparable across identity types — a stale admin on three platforms scores differently than a standard user with one low-severity finding.
Integration error classification¶
Integration errors are now classified by type: authentication expired, configuration invalid, rate limited, or API error. A sidebar badge shows the count of integrations needing attention, and a dismissible app-wide banner appears when critical auth errors require reconnection. No more guessing why a sync failed.
Trial extension + compliance preview¶
Free-tier users can now self-serve a one-time trial extension from the billing page. The Compliance page is visible in preview mode for free users with a plan gate, so prospects can evaluate the value before upgrading.
Other highlights¶
- Blast radius orbit visualization — Entity detail blast radius now renders as an interactive orbit diagram
- Behavioral baseline accuracy — Directory login events excluded from baselines when a dedicated IDP is connected
- Finding category consolidation — "Configuration" category folded into "Access Risk" and "Identity Security"
- AI prompt injection hardening — Topic-scoping guardrails prevent manipulation of AI assistant responses
- Billable identity fix — Count now only includes IDP/directory users, not SaaS-only accounts
- MTTR fix — Auto-resolved findings excluded from mean time to resolution calculations