Sublime Security ASA and ADÉ (Autonomous Security Analyst and Autonomous Detection Engineer)
Vendor: Sublime Security · HQ: United States · Category: Security
Last verified · Added · 5 sources · How we verify
Visit official site (opens in new tab) Claim this listing
Vendor? Get your “Listed on Agent Index” badge
What it does
Sublime Security is an email security platform built around customisable detections. Sublime's docs describe ASA (Autonomous Security Analyst) as an AI agent that clears the abuse mailbox and triages messages that the Attack Score model flagged as suspicious, unknown or error. ASA chooses which analysis tools to call for each message: attachment detonation, link analysis and computer vision on landing pages, and sender behaviour profiles. It can run parallel sub-investigations and usually reaches a verdict in about a minute. In passive ('Analyst') mode it investigates and writes a report without touching mail. In active ('Autonomous') mode it applies the auto-remediation you set per verdict (quarantine, trash, move to spam or warning banner) and auto-reviews. Verdicts are malicious, spam, graymail, benign or unknown, and unknown always goes to a human. When ASA confirms a miss it can hand the message to ADÉ (Autonomous Detection Engineer), which writes and backtests new detection rules. Docs say you can approve coverage yourself or let ADÉ auto-activate detections that clear the precision bar you set (the docs describe a 99% preset). ASA sends only parsed message data to the model, never the raw email. It runs in Sublime Cloud and in self-hosted AWS deployments, including AWS GovCloud, and needs Sublime Enterprise. Registry: the Sublime Security Platform is TX-RAMP Level 2 (TX1903681). That is a platform-level certification, and the listing doesn't name ASA. Speed and false-positive figures are vendor claims. Enterprise pricing is not published. A free Core tier and a free EML analyzer exist.
Frameworks
No framework tags verified yet.
Vendor-claimed TX-RAMP
Government authorizations
Registry records checked 2026-10-18. Scope “listing” means the certified product is this listing or the government boundary it already documents; “platform-only” means another product line is certified and the vendor does not state this agent is inside that boundary.
| Program | Product | Status | Scope | Source |
|---|---|---|---|---|
| TX-RAMP | Sublime Security Platform · Level 2 · TX1903681 · expires 2029-07-05 | Level 2 | Vendor platform; agent scope not stated | Registry |
Pricing (as published)
Not published for Sublime Enterprise (required for ASA). Free Sublime Core tier and free EML Analyzer available.
Tasks
phishing triageemail securityalert triageremediationthreat detection
Sources
- docs.sublime.security/docs/asa-autonomous-security-analyst (primary)
- sublime.security/
- docs.sublime.security/docs/ade-autonomous-detection-engineer
- legal.sublime.security/sublime-legal-terms/sublime-privacy-policy-4ace7a48
- dir.texas.gov/resource-library-item/tx-ramp-certified-cloud-products
Last verified 18 Oct 2026. Framework tags indicate relevance, not certification; “vendor-claimed” items are as marketed by the vendor. Spotted an error? Submit a correction or email sales@heritageservices.online. See our methodology.
Use-case comparisons
- Email security and phishing-response AI agents — 13 verified listings (Triage of user-reported phishing, autonomous email threat remediation and AI-run phishing simulation)
Compare Sublime Security ASA and ADÉ (Autonomous Security Analyst and Autonomous Detection Engineer) with
-
-
-
-
- A
-
-
-
-
-
Alternatives
-
7AI agentic security platform — Boston-based agentic security platform where swarms of specialised AI agents categorise, investigate and reach a determination on every alert, escalating only what needs a person, with optional managed service (PLAID ELITE). -
Abnormal AI Security Mailbox (Abnormal AI for Government) — Autonomous agent for email security operations that triages user-reported emails, removes related malicious campaign messages across the organisation and replies to reporters; included in Abnormal AI for Government, which is FedRAMP and GovRAMP Moderate authorized. -
Adaptive Security Agentic Email Security and Phish Triage — API-based 'agentic email security' that re-analyses inbound mail after native filtering with ML and LLM reasoning and removes confirmed threats across every inbox automatically (configurable human-in-the-loop thresholds). AI Phish Triage classifies employee-reported emails and acts on them, feeding Adaptive's simulations and awareness training. -
Aikido Attack AI Pentesting — Belgian autonomous AI pentest: many agents map the app, test attack paths and re-exploit each finding to confirm it, with a same-day audit-grade report. - A Andesite Human-AI SOC (Azimuth for Government) — Agentic SOC platform with configurable agents for alert triage, enrichment, investigation and remediation under analyst oversight; its government offering is FedRAMP High authorized.
-
Attic IVON (Agentic MDR for Microsoft 365) — Dutch AI SOC agent for Microsoft 365 SMBs that triages, investigates and contains identity alerts across Defender, Sentinel and Entra ID with human sign-off.
Stay updated
We use your email only to send the newsletter. See our privacy policy.