---
id: "contrarian-it-ownership"
type: "contrarian-insight"
source_timestamps: ["§ Redefining Ownership"]
tags: ["governance", "organizational-design"]
related: ["concept-lob-ai-ownership", "claim-lob-ownership", "action-shift-ownership-to-lob"]
challenges: "The conventional view that complex technology and AI models must be governed and managed by centralized IT or Data Science departments."
sources: ["agentic"]
sourceVaultSlug: "hbr-seg-agentic"
originDay: 6
articleStem: "hbr-sig-58-agent-managers"
sourceUrl: "https://hbr.org/2026/02/to-thrive-in-the-ai-era-companies-need-agent-managers"
sourceTitle: "To Thrive in the AI Era, Companies Need Agent Managers"
---
# AI deployment should not be owned by IT or Data Science

## Contrarian Insight — AI deployment should not be owned by IT or Data Science

**Challenges:** the conventional view that complex technology and AI models must be governed by centralized IT or Data Science departments.

Conventionally, new technology — especially complex AI — is owned, governed, and managed by IT or centralized Data Science teams. The authors call this a **pre-agentic mindset**. Because AI agents now execute **core business workflows**, ownership and accountability must shift to **Line of Business (LOB)** leaders. See [[concept-lob-ai-ownership]] and [[claim-lob-ownership]]; operationalized via [[action-shift-ownership-to-lob]].

**Counter-perspective (enrichment):** PyramidCI places 'Agent Operations' inside CIO/CTO/Transformation orgs as a horizontal control plane; Rasa insists on centralized security/compliance/governance. A hybrid — **centralized tooling & governance + decentralized process ownership** — is more realistic for large regulated firms. The provocation is directionally valuable but likely overstated in its absolute form.


## Related across articles
- [[action-remove-it-bottlenecks]]
- [[action-form-joint-governance]]