diff --git a/testing/backend/unit/test_finding_intelligence_confidence.py b/testing/backend/unit/test_finding_intelligence_confidence.py
new file mode 100644
index 00000000..ca71cd10
--- /dev/null
+++ b/testing/backend/unit/test_finding_intelligence_confidence.py
@@ -0,0 +1,335 @@
+"""
+Unit tests for finding_intelligence confidence and severity helper functions.
+
+Covers:
+- _normalize_severity: maps severity strings to canonical labels
+- _severity_rank: returns comparable integer rank for severity
+- _source_quality: scores source reliability from SOURCE_QUALITY map
+- _fingerprint_score: returns (score, match_strength) from finding metadata
+- _build_confidence_reason: composes human-readable confidence explanation
+- _compute_confidence: combines multiple factors into a 0-1 confidence score
+- _sort_sources: sorts and deduplicates source strings
+"""
+
+from __future__ import annotations
+
+import pytest
+
+from backend.secuscan.finding_intelligence import (
+    _normalize_severity,
+    _severity_rank,
+    _source_quality,
+    _fingerprint_score,
+    _build_confidence_reason,
+    _compute_confidence,
+    _sort_sources,
+)
+
+
+# ---------------------------------------------------------------------------
+# _normalize_severity
+# ---------------------------------------------------------------------------
+
+class TestNormalizeSeverity:
+    def test_known_severities(self):
+        assert _normalize_severity("critical") == "critical"
+        assert _normalize_severity("high") == "high"
+        assert _normalize_severity("medium") == "medium"
+        assert _normalize_severity("moderate") == "medium"
+        assert _normalize_severity("low") == "low"
+        assert _normalize_severity("info") == "info"
+        assert _normalize_severity("informational") == "info"
+        assert _normalize_severity("note") == "info"
+
+    def test_case_insensitive(self):
+        assert _normalize_severity("CRITICAL") == "critical"
+        assert _normalize_severity("High") == "high"
+        assert _normalize_severity("MeDiUm") == "medium"
+
+    def test_unknown_defaults_to_info(self):
+        assert _normalize_severity("unknown") == "info"
+        assert _normalize_severity("") == "info"
+        assert _normalize_severity(None) == "info"
+
+
+# ---------------------------------------------------------------------------
+# _severity_rank
+# ---------------------------------------------------------------------------
+
+class TestSeverityRank:
+    def test_rank_order(self):
+        assert _severity_rank("critical") == 5
+        assert _severity_rank("high") == 4
+        assert _severity_rank("medium") == 3
+        assert _severity_rank("low") == 2
+        assert _severity_rank("info") == 1
+
+    def test_unknown_defaults_to_info_rank(self):
+        assert _severity_rank("unknown") == 1
+        assert _severity_rank("") == 1
+        assert _severity_rank(None) == 1
+
+    def test_rank_is_comparable(self):
+        assert _severity_rank("critical") > _severity_rank("high")
+        assert _severity_rank("high") > _severity_rank("medium")
+        assert _severity_rank("medium") > _severity_rank("low")
+        assert _severity_rank("low") > _severity_rank("info")
+
+
+# ---------------------------------------------------------------------------
+# _source_quality
+# ---------------------------------------------------------------------------
+
+class TestSourceQuality:
+    def test_known_sources(self):
+        assert _source_quality(["nuclei"]) == 0.8
+        assert _source_quality(["nikto"]) == 0.7
+        assert _source_quality(["nmap"]) == 0.78
+        assert _source_quality(["http_probe"]) == 0.82
+
+    def test_multiple_sources_returns_max(self):
+        assert _source_quality(["nikto", "nuclei"]) == 0.8
+        assert _source_quality(["crawl", "openapi"]) == 0.8
+
+    def test_unknown_source_uses_default(self):
+        assert _source_quality(["unknown_scanner"]) == 0.58
+
+    def test_empty_source_list(self):
+        assert _source_quality([]) == 0.58
+
+    def test_whitespace_source_uses_default(self):
+        assert _source_quality(["  "]) == 0.58
+
+
+# ---------------------------------------------------------------------------
+# _fingerprint_score
+# ---------------------------------------------------------------------------
+
+class TestFingerprintScore:
+    def test_validated_match(self):
+        score, strength = _fingerprint_score({"validated": True})
+        assert score == 1.0
+        assert strength == "validated"
+
+    def test_exact_match(self):
+        score, strength = _fingerprint_score(
+            {"metadata": {"match_strength": "exact"}}
+        )
+        assert score == 0.95
+        assert strength == "exact"
+
+    def test_strong_fuzzy(self):
+        score, strength = _fingerprint_score(
+            {"metadata": {"match_strength": "strong_fuzzy"}}
+        )
+        assert score == 0.8
+        assert strength == "strong_fuzzy"
+
+    def test_fuzzy(self):
+        score, strength = _fingerprint_score(
+            {"metadata": {"match_strength": "fuzzy"}}
+        )
+        assert score == 0.7
+        assert strength == "fuzzy"
+
+    def test_family(self):
+        score, strength = _fingerprint_score(
+            {"metadata": {"match_strength": "family"}}
+        )
+        assert score == 0.45
+        assert strength == "family"
+
+    def test_none_match(self):
+        score, strength = _fingerprint_score({"metadata": {}})
+        assert score == 0.25
+        assert strength == "none"
+
+    def test_missing_metadata(self):
+        score, strength = _fingerprint_score({})
+        assert score == 0.25
+        assert strength == "none"
+
+
+# ---------------------------------------------------------------------------
+# _build_confidence_reason
+# ---------------------------------------------------------------------------
+
+class TestBuildConfidenceReason:
+    def test_basic_reason(self):
+        result = _build_confidence_reason(
+            finding_kind="observation",
+            evidence_count=3,
+            corroborating_sources=["nuclei"],
+            occurrence_count=1,
+            match_strength="none",
+        )
+        assert "Observation" in result
+        assert "3 evidence items" in result
+        assert "1 source" in result
+        assert result.endswith(".")
+
+    def test_singular_evidence(self):
+        result = _build_confidence_reason(
+            finding_kind="suspected_issue",
+            evidence_count=1,
+            corroborating_sources=[],
+            occurrence_count=1,
+            match_strength="none",
+        )
+        assert "evidence item" in result
+
+    def test_multiple_occurrences(self):
+        result = _build_confidence_reason(
+            finding_kind="validated_issue",
+            evidence_count=5,
+            corroborating_sources=["nuclei", "nmap"],
+            occurrence_count=3,
+            match_strength="fuzzy",
+        )
+        assert "3 scan observations" in result
+
+    def test_fingerprint_match(self):
+        result = _build_confidence_reason(
+            finding_kind="suspected_issue",
+            evidence_count=2,
+            corroborating_sources=[],
+            occurrence_count=1,
+            match_strength="exact",
+        )
+        assert "exact fingerprint match" in result
+
+    def test_empty_match_strength(self):
+        result = _build_confidence_reason(
+            finding_kind="observation",
+            evidence_count=2,
+            corroborating_sources=[],
+            occurrence_count=1,
+            match_strength="",
+        )
+        assert result  # should not crash
+
+
+# ---------------------------------------------------------------------------
+# _compute_confidence
+# ---------------------------------------------------------------------------
+
+class TestComputeConfidence:
+    def test_returns_float_in_valid_range(self):
+        score = _compute_confidence(
+            {},
+            corroborating_sources=["nuclei"],
+            occurrence_count=1,
+            evidence=[],
+        )
+        assert 0.0 <= score <= 0.99
+        assert isinstance(score, float)
+
+    def test_high_value_sources_increase_score(self):
+        low = _compute_confidence(
+            {},
+            corroborating_sources=["nikto"],
+            occurrence_count=1,
+            evidence=[],
+        )
+        high = _compute_confidence(
+            {},
+            corroborating_sources=["openapi"],
+            occurrence_count=1,
+            evidence=[],
+        )
+        assert high > low
+
+    def test_more_evidence_increases_score(self):
+        no_evidence = _compute_confidence(
+            {},
+            corroborating_sources=[],
+            occurrence_count=1,
+            evidence=[],
+        )
+        with_evidence = _compute_confidence(
+            {},
+            corroborating_sources=[],
+            occurrence_count=1,
+            evidence=[{}, {}, {}, {}],
+        )
+        assert with_evidence > no_evidence
+
+    def test_more_occurrences_increases_score(self):
+        single = _compute_confidence(
+            {},
+            corroborating_sources=[],
+            occurrence_count=1,
+            evidence=[],
+        )
+        repeated = _compute_confidence(
+            {},
+            corroborating_sources=[],
+            occurrence_count=5,
+            evidence=[],
+        )
+        assert repeated > single
+
+    def test_validated_finding_increases_score(self):
+        unvalidated = _compute_confidence(
+            {"validated": False},
+            corroborating_sources=[],
+            occurrence_count=1,
+            evidence=[],
+        )
+        validated = _compute_confidence(
+            {"validated": True},
+            corroborating_sources=[],
+            occurrence_count=1,
+            evidence=[],
+        )
+        assert validated > unvalidated
+
+    def test_critical_severity_increases_score(self):
+        info = _compute_confidence(
+            {"severity": "info"},
+            corroborating_sources=[],
+            occurrence_count=1,
+            evidence=[],
+        )
+        critical = _compute_confidence(
+            {"severity": "critical"},
+            corroborating_sources=[],
+            occurrence_count=1,
+            evidence=[],
+        )
+        assert critical > info
+
+    def test_score_is_capped_at_0_99(self):
+        # Very strong finding
+        score = _compute_confidence(
+            {"validated": True, "severity": "critical"},
+            corroborating_sources=["openapi", "graphql", "nuclei"],
+            occurrence_count=10,
+            evidence=[{}, {}, {}, {}],
+        )
+        assert score <= 0.99
+
+
+# ---------------------------------------------------------------------------
+# _sort_sources
+# ---------------------------------------------------------------------------
+
+class TestSortSources:
+    def test_removes_duplicates(self):
+        result = _sort_sources(["nuclei", "nuclei", "nikto"])
+        assert result == ["nikto", "nuclei"]
+
+    def test_sorts_alphabetically(self):
+        result = _sort_sources(["zap", "nuclei", "nikto"])
+        assert result == ["nikto", "nuclei", "zap"]
+
+    def test_whitespace_is_stripped(self):
+        result = _sort_sources(["  nuclei  ", "nikto"])
+        assert "nuclei" in result
+        assert "  nuclei  " not in result
+
+    def test_empty_and_whitespace_removed(self):
+        result = _sort_sources(["nuclei", "", "  ", None])
+        assert "nuclei" in result
+        assert "" not in result
+        assert "  " not in result