The Mobile World Congress 2026 in Barcelona earlier in March featured extensive discussions among telco leaders on AI’s transition from specialized uses to becoming essential for core network operations and digital services. But the most pressing concern highlighted was the rise of AI-driven threats, particularly digital fraud and cybercrime.
Telcos are crucial gatekeepers in the cyber ecosystem. Almost every scam, whether via phone calls, SMS or digital authentication, passes through their network infrastructure at some point. Historically, this has been the norm, but the AI era has introduced a new level of sophistication and speed of attacks, outpacing the evolution of traditional defenses.
The fraud economy
According to the Global Anti-Scam Alliance (GASA) and the World Economic Forum, scams cost victims an estimated $442 billion globally in 2025 alone. Approximately 50% of individuals report being targeted by scams at least once per week, while only 0.05% of cybercriminals are ultimately prosecuted.
Deepfake voice impersonation, highly personalized phishing messages, and automated social engineering campaigns can now be created at scale with minimal effort and cost. This highlights a stark imbalance: the economics of cybercrime currently favors attackers.
So, how can leaders mitigate the threats and continue to innovate while also preserving trust with their customers? This is a significant challenge, considering the industry’s historical emphasis on its reputation for network reliability, service availability and data security. AI has expanded the scope of what must now be trusted; from automated decision-making to authentication, fraud prevention to billing validation, and service prioritization to customer interactions.
When innovation meets system risks
Telecom networks underpin financial systems, healthcare infrastructure, public safety communications and industrial automation. At the enterprise level, organizations are using AI to power functions like:
- Real-time network routing and optimization
- Automated authentication and identity verification
- Fraud detection and prevention
- Predictive maintenance for network infrastructure
- Intelligent service orchestration
Protiviti’s AI Pulse Survey research found that AI adoption among telcos is steadily moving from pilots to optimization, with the most benefits concentrated where integration and governance are strongest. According to the research, telco leaders also expect to deploy agents primarily as semi‑autonomous assistants at first (bounded decision/action), with some organizations targeting fully autonomous agents as capabilities and controls mature.
Regardless of the autonomy level pursued, when AI systems perform well, the benefits are substantial. However, when AI systems fail, the consequences can scale just as quickly.
According to Protiviti’s research, telcos consistently framed fraud detection, risk management, and cyber related threats as priority use cases for AI and emerging agentic AI capabilities. The data, and by agreement the focus of the wide array of partners and technologies showcased at the Barcelona conference, showed a clear pattern: the sector is moving toward proactive detection, automation driven investigation support, and integration with enterprise risk systems.
There are some crucial challenges that must be addressed first to perfect these systems. A flawed fraud detection model could block legitimate users across millions of accounts. A poorly governed algorithm could introduce regulatory exposure. A compromised AI pipeline could create vulnerabilities across distributed network infrastructure.
To be successful, telco leaders must ask critical governance questions like:
- Who owns AI model risk across the enterprise?
- How are AI models validated before deployment and continuously monitored in production?
- How is bias, drift and performance degradation identified and addressed?
- What level of human oversight exists for high-impact automated decisions?
These are not purely technical considerations. They are governance questions that affect regulatory exposure, brand reputation, and long-term customer trust.
Trust: a competitive differentiator
In markets where network performance differences are narrowing, trust is becoming an important differentiator. Consumers are increasingly aware of digital risks. Enterprise customers deploying AI workloads, IoT environments, and mission-critical services expect secure and resilient connectivity.
Operators that show responsible AI governance, proactive fraud prevention, and resilient infrastructure will be better positioned to retain customers and attract enterprise partners. In enterprise markets in particular—where connectivity supports AI platforms, industrial automation, and intelligent edge services—confidence in an operator’s ability to secure and govern intelligent systems can directly influence vendor selection.
Trust influences customer loyalty, contract renewals and the adoption of premium digital services.
Engineering (in) trust
Trust cannot be assumed simply because AI improves efficiency. It must be deliberately engineered; operationalized across technology, governance and collaboration. These three pillars are emerging as foundational to this effort:
- Infrastructure reliability: AI embedded in telecom operations must meet infrastructure-grade expectations. This includes secure data pipelines, resilient model deployment environments, continuous monitoring and adversarial testing against manipulation attempts. In many environments, AI now influences network performance, authentication decisions and financial transactions. It must therefore be treated as part of the critical infrastructure stack.
- Transparent governance: Explainability and auditability are increasingly essential. Organizations should ensure models are documented, version-controlled, and subject to structured validation processes. Human oversight remains important for decisions affecting customer access, billing and identity verification.
- Ecosystem collaboration: Fraud and digital abuse cross both industry and national boundaries. Effective defense requires intelligence sharing, common fraud signals, and coordinated security frameworks across telecom operators, financial institutions and digital platforms. Operators that actively participate in collaborative security ecosystems strengthen both industry resilience and their own market credibility.
What telecom operators can do now
Here are several practical steps telecom leaders can take to build trust into their AI-enabled environments from the outset.
- Treat AI as critical infrastructure: It’s important that organizations govern with the same rigor as core telecom infrastructure. This includes resilience, security, and performance standards aligned to mission-critical systems.
- Establish clear ownership of AI risk: Define accountability for AI use at the enterprise level, including ownership of model risk, governance, and performance. This ensures AI decisions are subject to the same oversight as other critical components.
- Implement continuous model validation and monitoring: Move beyond one-time validation by establishing processes to monitor model performance, drift, and unintended outcomes in production. AI infrastructure should be continuously evaluated as conditions and data evolve.
- Integrate fraud prevention into AI strategy: Fraud detection should not operate as a standalone function. Operators should embed fraud prevention capabilities directly into AI-driven network and customer interaction workflows.
- Strengthen governance for automated systems and identities: As AI agents and automated processes increase, establish controls to govern how these systems authenticate, access data, and interact across environments.
- Participate in cross-industry collaboration: Fraud and digital threats extend beyond telecom boundaries. Engaging in intelligence sharing and coordinated defense efforts strengthens both ecosystem resilience and individual operator trust posture.
As telecom networks continue to advance in intelligence, the intersection of AI governance, cybersecurity, resilience, and fraud prevention is becoming increasingly important.
At Protiviti, our teams are working with telecom operators to develop and implement strategies that manage AI integration with the same rigor as critical infrastructure. We believe that going forward, the ability to operate AI-led infrastructure that is trusted at scale will define how telecom operators compete and grow successfully.
