Scores and prioritizes risk from data to support faster, more consistent decisions.
Use cases
100
Examples
60
Industries
12
Timeline
19 mo
Data updated 1 day ago
Adoption over time
Documented cases per month
By case publish month · completed months only
79 cases documented across 37 months (Jul 23 – Jul 26), peaking at 11 in April 2025.
AI Use Cases Hub
10 earlier cases before Jul 23 not shown
Each column counts every documented case of this type by its publish month, across the full corpus. The in-progress current month is excluded from columns and surfaced separately, and cases published before the charted window are summarized as earlier cases instead of plotted.
3.4Innovativeness3.4/5Differentiated3.4/5 - Differentiated. Comparable to recent managed-service AI/security cases around the 3.x range: it combines AI algorithms with multiple Alibaba Cloud data and analytics services for a specialized Web3 security workflow, but the article does not show a novel model architecture or breakthrough scale.
Trusta Labs is a Web3 infrastructure company in Hong Kong that builds security and data infrastructure through AI.It uses Alibaba Cloud elastic computing, storage, data management, and analytics services to match on-chain assets with the right users and to support AI-powered Sybil attack prevention, knowledge graph-enhanced profiling, and user targeting.
2.9Innovativeness2.9/5Differentiated2.9/5 - Differentiated. A practical cloud migration and analytics modernization for scenario risk modelling; the article shows useful orchestration across BigQuery, Dataflow, and Cloud Storage, but not a particularly novel AI architecture.
HSBC built a scenario risk-modelling Risk Advisory Tool on Google Cloud to move high-volume scenario analysis and risk modelling off limited on-premises hardware and accelerate decision support for traders and risk-management teams.The implementation used Google Cloud services including Cloud Storage, Dataflow, and BigQuery.
4.1Innovativeness4.1/5Advanced4.1/5 - Advanced. This is an advanced insurance underwriting stack because it combines geospatial modeling, AI/ML pipelines, and real-time quote generation in a market where risk is highly dynamic. Compared with recent similar cloud AI cases, it is notably more specialized and operationally consequential than a common assistant or basic RAG workflow.
Delos Insurance Solutions needed to make wildfire-risk underwriting viable for California homeowners who were often considered uninsurable by other carriers.The company built wildfire-behavior and geospatial risk models with Google Earth Engine, then used Colab notebooks and Vertex AI pipelines for model training and deployment, plus Google Maps Platform for a quote-to-buy underwriting portal.The solution let Delos price policies more accurately, offer coverage in high-risk areas, and support agents with a fast online underwriting workflow.
3.3Innovativeness3.3/5Differentiated3.3/5 - Differentiated. Uses a managed foundation model platform with data consolidation and fraud screening; more differentiated than a basic chatbot, but not an uncommon architecture compared with recent Amazon Bedrock enterprise cases.
KOHO rebuilt credit infrastructure to assess credit-invisible customers using behavioral data such as rent, utility, and phone payment records.The company built Kortex AI on Amazon Bedrock to extract behavioral insight for underwriting and used AI screen detection to reduce account takeovers.
2.7Innovativeness2.7/5Differentiated2.7/5 - Differentiated. This is an operationally mature but not novel architecture: the distinctive element is adopting Google’s SRE model for public-sector cybersecurity operations, which is more implementation/process innovation than breakthrough AI or platform innovation. Relative to recent Google Cloud public-sector cases, it is closer to a differentiated infrastructure-operations pattern than to advanced AI cases.
New York City Cyber Command (NYC3) built a security log aggregation platform on Google Cloud to keep city security systems available 24/7 and support risk alerts, visualization, and analytics for 100+ city agencies.The organization adapted Google’s Site Reliability Engineering model, created an internal SRE team, and trained/cross-trained engineers across development, networking, and operations to improve reliability and supportability.NYC3’s internally developed Juggernaut threat-management system parses real-time cybersecurity data from disparate sources and alerts the team when something seems amiss.
3Innovativeness3/5Differentiated3/5 - Differentiated. This is a solid applied Google Cloud analytics modernization for risk modeling, but it is close to other recent BigQuery/analytics transformation cases and does not show a particularly novel AI architecture.
Global bank HSBC launched a new scenario risk-modelling tool on Google Cloud for the risk management and trading teams.The tool supports intraday, self-service scenario analysis and helps users assess portfolio risk, capital requirements, and default-risk exposure.
4Innovativeness4/5Advanced4/5 - Advanced. Advanced because it combines production agentic investigation, expert-validation at scale, and transparent reasoning for security operations; compared with recent 3-to-4 point agent cases, this one is stronger on validated decision quality and operational security constraints.
eSentire used Anthropic Claude in Amazon Bedrock to augment its managed detection and response security operations.The system formulates investigation hypotheses from threat indicators, dynamically selects and executes evidence-gathering tools, and produces interactive investigation reports with evidence and reasoning chains.The company validated model outputs against senior SOC experts and uses AWS services including Lambda, API Gateway, IAM, and CloudWatch to support the production workflow.
3Innovativeness3/5Differentiated3/5 - Differentiated. This is a pragmatic generative AI deployment for surveillance and fraud/AML workflows, similar in novelty to recent AWS financial-services cases and not showing a notably advanced architecture beyond applied GenAI/coplilot-style capabilities.
Nasdaq implemented AWS generative AI to improve market surveillance for regulators and marketplaces globally.The solution streamlines triage and examination, helping clients more effectively monitor and detect potential market manipulation and insider dealing.The source also describes additional capital-markets applications for anti-money laundering, fraud prevention, and Entity Research Copilot workflows at Nasdaq Verafin.
3.4Innovativeness3.4/5Differentiated3.4/5 - Differentiated. Similar to other managed-AWS agentic AI cases, but slightly more advanced because it combines near-real-time security decisioning at very large event volume with Bedrock-powered agents integrated into an existing orchestration framework.
Human Managed built its I.DE.A. platform to unify intelligence, decisions, and actions across more than 40 telemetry sources into a single framework for near-real-time risk and security decisions.The company is expanding AWS usage to integrate Amazon Bedrock into its MCP framework for enterprise generative AI agents that support summarization, contextual querying, cross-source analysis, decision support, and risk/compliance automation.
3Innovativeness3/5Differentiated3/5 - Differentiated. Compared with recent calibration cases, this is a solid but common governed analytics/self-service reporting implementation rather than an advanced AI architecture; the main novelty is operational efficiency at scale, not a novel model or agent workflow.
Landbay is a UK financial technology company that operates a peer-to-peer lending platform for buy-to-let mortgages.The company replaced spreadsheet-based reporting with Looker on Google Cloud to give internal teams, auditors, and clients trusted self-service access to consistent data.Looker supports real-time portfolio stratification and automatically generated lender-specific reports, helping Landbay scale data-driven decision-making and underwriting oversight.
How many risk assessment use cases are documented?
The AI Use Case Hub documents 100 real risk assessment deployments across 12 industries, with 60 detailed company examples you can browse.
Which industries adopt risk assessment the most?
Risk assessment is most common in Finance (47%), Insurance (14%) and Professional Services (9%).
Which countries lead in risk assessment?
United States leads documented risk assessment deployments, followed by United Kingdom and Global.
What technologies are used for risk assessment?
Teams most often build risk assessment with Azure OpenAI, Azure AI and Azure.
What AI capabilities power risk assessment?
Across the documented deployments, the most common capability patterns are Agent (22%), Copilot (20%) and Multi-agent (7%).
What results do companies report from risk assessment?
Across the 100 deployments reporting outcomes, companies most often cite risk & compliance (68%), speed & agility (62%) and new product / capability (47%). Where impact is quantified, the strongest evidence is in time & speed: a median −47.5% across 8 reported metrics.