Cisco's AI-Driven Voice Security Cuts Fraud and Boosts Efficiency
Discover how Cisco IT leveraged AI to modernize its voice security, achieving significant reductions in toll fraud and manual investigations, while addressing critical compliance liabilities.
Key Facts
- Cisco achieved a 70% reduction in toll fraud, highlighting the financial impact of AI on security.
- A 60% decrease in manual investigations reveals operational efficiency and reduced labor costs.
- The shift to predictive security shows a strategic move from reactive to proactive threat management.
- Integration with Splunk Cloud enables real-time data analysis, enhancing competitive positioning in security.
- Partnering with CX teams accelerates innovation, indicating a strategic focus on customer-centric solutions.
Summary
Summary
Cisco IT faced significant challenges with increasing toll fraud and nuisance calls, which burdened their security teams and exposed them to compliance risks. They implemented a hybrid AI-driven solution that combined machine learning with traditional security measures, resulting in a 70% reduction in toll fraud and a 60% decrease in manual investigation efforts.
Background
Cisco IT operates one of the largest enterprise voice environments globally, managing vast volumes of calls daily. Before deploying their AI solution, they relied on outdated methods such as manual blocklists and static controls, which were inadequate against the evolving threat landscape of robocalls and toll fraud.
Challenge
The primary issue was the inefficiency of traditional defenses that could not keep pace with the sophistication of threats. The existing systems provided limited context, requiring human intervention for each investigation, which created delays and left Cisco vulnerable to fraud and regulatory scrutiny.
Solution
Cisco built a composite risk-scoring engine that integrated their on-prem Cisco Unified Communications infrastructure with the cloud-native Webex Calling environment. They utilized the Splunk Cloud Platform to normalize and analyze call data in real-time, enabling them to correlate voice telemetry with broader security data. This system employed machine learning models, including Random Forest and XGBoost, to evaluate calls based on enriched features and external threat intelligence.
Results
The deployment led to a 60% reduction in manual investigation efforts and an estimated 70% decrease in potential toll fraud losses. Additionally, the system improved the speed of fraud detection and significantly reduced the volume of spam calls reaching employees, thereby restoring trust in Cisco's voice services.
Key Insights
Organizations should prioritize visibility in their security measures, moving beyond static rules to incorporate behavioral intelligence. This approach is scalable and can be adapted to various infrastructures, whether through internal development or leveraging third-party solutions.
Customer Testimonial
No direct quote is available from the source material.
Entities Mentioned
Companies
Products
Technologies
Organizations
Key Concepts
Definitions
- AIOps
- AIOps refers to the application of artificial intelligence to IT operations, enabling organizations to analyze large volumes of data and automate processes.
- observability
- Observability is the ability to measure and understand the internal state of a system based on the data it generates.
- toll fraud
- Toll fraud is a type of telecommunications fraud where unauthorized calls are made, leading to financial losses for organizations.
- machine learning
- Machine learning is a subset of artificial intelligence that enables systems to learn from data and improve their performance over time without being explicitly programmed.
- compliance liability
- Compliance liability refers to the risk of legal penalties or fines due to failure to adhere to regulatory requirements.
Use Cases
- →reducing toll fraud losses
- →automating call investigation
- →enhancing voice service security
- →improving operational efficiency
- →restoring trust in voice communications
- →leveraging AI for threat detection
Frequently Asked Questions
How did Cisco IT modernize its voice security?
Cisco IT modernized its voice security by applying AI and observability principles to better understand call data and mitigate threats proactively. This approach led to significant reductions in toll fraud and manual investigation efforts.
What technologies were used in the new voice security solution?
The new voice security solution utilized AIOps, machine learning models like Random Forest and XGBoost, and the Splunk Cloud Platform to aggregate and analyze call data in real-time.
What were the results of implementing the new voice security measures?
The implementation resulted in a 70% reduction in potential toll fraud losses and a 60% reduction in manual investigation efforts, significantly enhancing operational efficiency and security.
What is the role of Cisco's Customer Experience team in this project?
Cisco's Customer Experience team played a crucial role by combining operational scale with specialized expertise, accelerating the development cycle and ensuring the solution met rigorous standards for global client deployments.
What advice does Cisco give to organizations looking to modernize voice security?
Cisco advises organizations to start with visibility, moving away from static rules and applying behavioral intelligence to effectively measure and address voice security challenges.