-
Trust Insightsについて
SNS詐欺や脅迫による被害を防ぐTrust Insightsについて解説します。この新しいフレームワークでは、プライバシー保護のための機械学習を利用し、ユーザーが高リスクな行動に誘導される可能性がある状況を検出します。Trust Insightsをアプリに統合し、シグナルを解釈して、プライバシーを尊重しながらユーザーを保護する適切な介入策を設計する方法を確認しましょう。
関連する章
- 0:00 - Introduction
- 2:35 - Generating insights
- 6:50 - Feedback requirements
- 9:25 - Privacy
- 10:34 - Best practices
- 12:48 - Next steps
リソース
関連ビデオ
WWDC26
-
このビデオを検索
-
-
3:01 - Generating insights
import TrustInsights let request = IsLikelyBeingCoachedInsight.request(schema: .version1, modelVersion: .current) let context = InsightEvaluator.InsightContext(operationCategory: .resourceUse, requestedEvaluations: request) let evaluator = InsightEvaluator() guard try await evaluator.requestAuthorization(for: context) == .authorized else { return } let assessment = try await evaluator.requestEvaluation(context: context) do { try handleAssessment(assessment) } catch { // Handle error } assessment.reportConsumption(.usedIncreasedFriction) -
5:37 - Handling results for IsLikelyBeingCoachedInsight
func handleAssessment(_ assessment: InsightEvaluation<IsLikelyBeingCoachedInsight>) throws { switch try assessment.insight.outcome.get() { case .unknown: case .medium: case .high: @unknown default: } } -
7:05 - Real-time consumption feedback
import TrustInsights let request = IsLikelyBeingCoachedInsight.request(schema: .version1, modelVersion: .current) let context = InsightEvaluator.InsightContext(operationCategory: .resourceUse, requestedEvaluations: request) let evaluator = InsightEvaluator() guard try await evaluator.requestAuthorization(for: context) == .authorized else { return } let assessment = try await evaluator.requestEvaluation(context: context) do { try handleAssessment(assessment) } catch { // Handle error } assessment.reportConsumption(.usedIncreasedFriction)
-
-
- 0:00 - Introduction
Meet Trust Insights, a new iOS 27 framework that helps your app detect coercion and social engineering.
- 2:35 - Generating insights
Integrating Trust Insights with its client-side Swift API — declaring the entitlement, building a parameter pack of requested insights, and using the InsightEvaluator with an operation category that determines which model logic applies.
- 6:50 - Feedback requirements
The two required types of feedback that keep insights accurate — mandatory real-time consumption feedback reporting how your app responded, and offline feedback for transactions that later prove fraudulent.
- 9:25 - Privacy
Understand how Trust Insights minimizes data, keeps signals on device, and gives users full control.
- 10:34 - Best practices
Learn where Trust Insights adds the most value and how to combine it with your existing risk logic.
- 12:48 - Next steps
Adopting Trust Insights — identify moments where it can work alongside your existing logic, follow the documentation and best practices, and register on Apple Business Register to learn about Partner Data Services.