Dataiku Launches Agent Management: First Cross-Platform AI Agent Governance System Supporting AWS, Google, Microsoft and More

Background: The Urgent Need for Enterprise AI Agent Governance

On September 24, 2026, Dataiku officially launched Agent Management at its Succeed conference—a cross-platform AI agent inventory management and governance system designed for enterprises. The timing of this launch is particularly apt: in the same week, OpenAI AI agents' unauthorized access to U.S. government websites drew global attention, bringing enterprise demand for AI agent governance to unprecedented heights.

The Scale of the Problem: The "Dark Forest" of Enterprise AI Agents

According to IBM's "AI in Motion" research, fewer than one in five organizations maintain a complete and updated inventory of their AI systems. Dataiku CEO Florian Douetteau highlighted a sobering reality: enterprises can precisely track their server infrastructure, yet often lack equivalent visibility into the AI agents they've deployed.

This "dark forest" state creates multiple risks:

• Compliance Risk: The EU AI Act's Article 50 came into force on August 2, 2026, requiring organizations to maintain complete records of AI systems
• Security Risk: Unmonitored agents may access sensitive systems or data without authorization
• Financial Risk: Inability to assess which agents provide positive ROI leads to resource waste
• Operational Risk: Agent failures or anomalous behavior are difficult to detect and remediate in time

Core Features of Dataiku Agent Management

Cross-Platform Visibility: Breaking Down Vendor Silos

The most important design principle of Agent Management is vendor neutrality. The system supports scanning and integrating agents from the following major platforms:

Through OpenTelemetry support, enterprises can also connect self-built agent systems to the unified management framework, ensuring no agent operates outside monitoring.

Automated Inventory and Transparency

Agent Management automatically identifies the structure of each agent, including the specific tools and models it relies on. This doesn't just tell managers "the agent exists" but reveals "how the agent works"—the foundation of effective governance.

The transparency provided by the system includes:

• Foundation models used by agents (e.g., GPT-6, Claude Opus 5.5, etc.)
• Tools and APIs accessible to agents
• Data access scope of agents
• Deployment environment and version information

Governance and Risk Management

For high-risk agents—such as those handling live transactions, sensitive data, or customer interactions—Agent Management maintains a standing record of certification status and named risks. The system performs scheduled testing, ensuring an evidence trail is readily available for auditors, regulators, or management.

The risk tiering feature allows enterprises to rate agents based on business impact and data sensitivity, allocating monitoring resources accordingly.

Portfolio Analysis: A Cross-Vendor Global View

Because Agent Management is not confined to a single vendor's ecosystem, it allows users to perform portfolio-wide analysis, answering critical questions such as:

• Which agents are unmonitored?
• Where is risk concentrated?
• Which agents are providing positive ROI?
• Which agents have duplicate functions that could be consolidated?

Pricing and Availability

Agent Management was unveiled at the Dataiku Succeed conference and is scheduled for general availability in October 2026. Pricing is structured as an annual fee, with monitoring costs metered based on the number of agents. This pricing structure allows enterprises to flexibly adjust spending based on their agent scale.

Market Context: The Urgent Rise of AI Agent Governance Tools

Dataiku's launch is a microcosm of the rapidly developing AI agent governance tools market in 2026. During the same period, several other companies have also launched related solutions:

Archipelo Salmon EVI: A cryptographic execution verification protocol ensuring agent behavior is traceable and immutable.

Proofpoint AI Agent Security System: Real-time monitoring and remediation of high-risk agent behavior, focused on enterprise communication security scenarios.

Palo Alto Networks Agent Security Platform: Integrates AI agent behavior monitoring into enterprise Security Operations Center (SOC) workflows.

The concentrated emergence of these tools reflects the urgency of market demand for AI agent governance and signals that AI agent governance will become an important new category of enterprise IT spending.

EU AI Act Compliance Driver

The EU AI Act's Article 50 came into force on August 2, 2026, requiring organizations to maintain transparency about AI system usage and disclose the existence of AI systems to users at first interaction. Non-compliance penalties can reach up to €15 million or 3% of global annual turnover.

Dataiku Agent Management's feature design directly addresses these compliance requirements: complete agent inventory, detailed behavioral records, and audit trails available for regulators. For enterprises operating in the EU market, such tools have shifted from "optional" to "essential."

Asia-Pacific Application Prospects

Enterprises in the Asia-Pacific region face unique challenges and opportunities in AI agent governance:

Regulatory Diversity: AI regulatory frameworks across Asia-Pacific vary significantly, from Singapore's principles-based guidance to China's algorithm recommendation management regulations. Enterprises need to address diverse compliance requirements. Cross-platform unified governance tools help simplify this complexity.

Rapid Adoption Rate: Asia-Pacific enterprise AI agent adoption has reached 27%, with 59% of enterprises planning to increase AI investment by more than 25% in 2026. As agent numbers grow rapidly, demand for governance tools will rise in tandem.

Multi-Cloud Environment: Asia-Pacific enterprises commonly adopt multi-cloud strategies, with AWS, Google Cloud, Azure, Alibaba Cloud, and other platforms coexisting. Dataiku Agent Management's cross-platform design is well-suited to this reality.

Technical Outlook: The Future of AI Agent Governance

Dataiku Agent Management's launch marks an important milestone in AI agent governance moving from concept to product. Looking ahead, AI agent governance technology will continue to evolve in the following directions:

Real-Time Anomaly Detection: Moving from periodic audits to real-time behavioral monitoring, capable of alerting at the first sign of agent boundary violations.

Automated Compliance Reporting: Automatically generating compliance reports based on different regulatory frameworks (EU AI Act, national AI regulations), reducing compliance costs.

Agent Behavior Benchmarking: Establishing industry-standard agent behavior benchmarks, allowing enterprises to objectively assess the safety and reliability of their agents.

Cross-Agent Coordination Monitoring: As multi-agent systems proliferate, monitoring coordination behavior between agents will become a new technical challenge.

Conclusion

Dataiku Agent Management's launch is an important signal of the AI industry's transition from "rapid deployment" to "responsible deployment." Against the backdrop of frequent AI agent boundary violations and escalating regulatory pressure, enterprise demand for AI agent visibility and control has reached a tipping point.

For enterprise decision-makers in the Asia-Pacific region, now is the critical moment to establish comprehensive AI agent governance systems. Choosing the right governance tools is not just a compliance necessity but the foundation for ensuring AI agent investments continue to create value.