Splunk Launches Agentic AI Observability Platform, Targeting European Banks and Telcos

Traditional observability platforms collect logs, metrics, and traces, then surface dashboards and alerts for human operators to interpret. The model works, but it has a ceiling. As distributed systems grow more complex, the volume of signals outpaces human capacity to triage them. Alert fatigue is endemic across enterprise IT operations teams throughout the EU and UK.

Splunk's agentic AI approach addresses this directly. Rather than simply flagging an anomaly, the AI agents are designed to reason across multiple data streams simultaneously, identify probable root causes, and propose or enact fixes. This mirrors what a skilled site reliability engineer would do, but at machine speed and at a scale no human team can match.

The agents can also correlate events across siloed systems that historically required manual cross-referencing. In a typical enterprise environment, a degraded customer experience might trace back to a network misconfiguration, a microservices timeout, and a database latency spike occurring in parallel. Identifying that chain manually is slow and error-prone. Agentic AI is built precisely for that kind of multi-variable reasoning.

The Competitive Landscape for AI-Powered Observability

Splunk is not moving in isolation. The AI observability space is crowded and accelerating. Datadog has been aggressively expanding its AI features, including its Bits AI assistant and automated investigation workflows. New Relic has similarly pushed AI-assisted analysis into its core platform. Both are well-funded, cloud-native competitors with strong developer mindshare across European enterprise accounts.

What Splunk brings that neither fully replicates is depth of integration with Cisco's broader security and networking portfolio. For enterprise customers already invested in Cisco infrastructure, the ability to correlate observability data with network telemetry and security signals from the same vendor is a genuine differentiator. The acquisition thesis was always about creating a unified data platform across security operations and IT operations, and the agentic AI rollout is the first clear expression of that combined capability.

The European Regulatory Picture

European financial services firms and telecoms operators are precisely the organisations where system downtime carries regulatory, financial, and reputational consequences that make faster, autonomous remediation genuinely valuable. The EU's Digital Operational Resilience Act (DORA), which came into full effect on 17/01/2025, mandates that financial entities demonstrate robust ICT risk management and rapid incident response. Agentic AI that can autonomously detect and begin remediating issues fits directly into this compliance posture, provided the audit trail and governance controls are rigorous enough to satisfy supervisors at the European Banking Authority.

Lucilla Sioli, Director for Artificial Intelligence and Digital Industry at the European Commission, has consistently emphasised that AI systems deployed in critical infrastructure must be accompanied by robust human oversight mechanisms. That framing maps directly onto the challenge Splunk faces: persuading compliance teams at major European banks that autonomous remediation actions are sufficiently auditable and controllable to meet both DORA and the EU AI Act's requirements for high-risk AI system governance.

The UK's position post-Brexit adds a further dimension. The Financial Conduct Authority has published guidance indicating that firms deploying AI in operational contexts must maintain clear accountability chains. Gary Gensler's successor frameworks aside, the FCA's own 2024 AI update made clear that autonomous decision-making in regulated infrastructure is permissible only where explainability and human override capabilities are demonstrably in place. Splunk's governance controls will receive close scrutiny from UK financial institutions before wide-scale autonomous remediation is sanctioned in production.

Germany's major banks, including Deutsche Bank and Commerzbank, both of which have significant Cisco infrastructure investments, represent obvious early targets for the platform. The same applies to BT Group and Vodafone in the UK, both managing complex hybrid environments as 5G rollout continues alongside ageing legacy infrastructure. IT complexity in Europe's financial and telecoms sectors is growing faster than teams can scale, and autonomous observability is rapidly shifting from a desirable capability to an operational and regulatory necessity.

Implications for Enterprise IT Teams Across Europe

The shift to agentic observability does not eliminate the need for skilled engineers. It changes what those engineers do. The first-line triage and correlation work that currently consumes significant engineering hours becomes automated. Engineers are repositioned toward policy setting, exception handling, and strategic architecture decisions rather than reactive incident response.

This transition is not without friction. Organisations need to:

• Define clear boundaries for what AI agents are permitted to action autonomously versus what requires human approval
• Establish audit and logging standards to satisfy internal governance and external regulatory requirements, particularly under DORA and the EU AI Act
• Retrain operations teams to work alongside AI agents rather than replacing the human judgement layer entirely
• Validate that AI-generated remediation recommendations are tested against their specific environment configurations

There is also a broader workforce consideration. As platforms like Splunk automate more of the routine observability workload, the skills premium shifts toward engineers who understand how to configure, govern, and extend AI systems, rather than those who manually triage alerts. Research from ETH Zurich's AI Centre has documented this pattern across AI-augmented professional roles, noting that cognitive overload can emerge when teams transition too quickly without adequate change management structures in place.

For smaller IT teams, particularly at mid-market firms across the EU's industrial heartlands in Germany, France, and the Netherlands, the autonomous capability may be especially valuable. Organisations without 24-hour staffing models gain round-the-clock monitoring and initial response without proportional headcount costs. That is a concrete competitive advantage, not a theoretical one.

What Comes Next

The agentic AI rollout is an early-stage deployment and the full capability set will mature over successive releases. Key areas to watch include how deeply the agents can integrate with Cisco's security operations tools, specifically whether observability and security telemetry can be correlated by the same AI reasoning layer. That would represent a meaningful step toward the unified platform Cisco articulated when it announced the acquisition.

Competitor responses will accelerate. Datadog and New Relic will push their own agentic features, and newer entrants building AI-native observability from the ground up, including European players such as Dynatrace, which is already positioning its Davis AI engine as an autonomous causation engine, will continue to apply pressure. The next 12 to 18 months in enterprise observability tooling will be defined by how effectively platforms can demonstrate that autonomous AI agents reduce mean time to resolution and engineering burden in production environments, not just in controlled demonstrations. European enterprises should treat 2025 as the year to run structured pilots, not the year to commit wholesale.