---
id: "claim-software-engineering-agentic-activity"
type: "claim"
source_timestamps: ["¶3"]
tags: ["statistics", "software-engineering", "ai-adoption"]
related: ["claim-marketing-bottleneck", "entity-anthropic"]
confidence: "high"
external_confidence: "low — not independently verifiable"
testable: true
speakers: ["Michelle Taite", "John Winsor", "Will Fernandez"]
sources: ["agentic"]
sourceVaultSlug: "hbr-seg-agentic"
originDay: 6
articleStem: "hbr-tier1-02-agentic-marketing-org"
sourceUrl: "https://hbr.org/2026/05/redesigning-your-marketing-organization-for-the-agentic-age"
sourceTitle: "Redesigning Your Marketing Organization for the Agentic Age"
---
# Software Engineering Accounts for Nearly 50% of Agentic Activity

**Claim:** Citing research from [[entity-anthropic-d6]], the authors state that software engineering now accounts for **nearly 50% of all agentic AI activity** in the enterprise.

This massive concentration of AI adoption in engineering is the primary driver behind faster product-development cycles, which in turn places unprecedented strain on marketing departments that have not adopted agentic models at the same pace (see [[claim-marketing-bottleneck]]).

**Confidence:** High as author-stated · **Testable:** Yes.

**Validation (enrichment) — treat with caution:** *Not directly verifiable.* Open web search does **not** surface a specific, widely-cited Anthropic statistic that "software engineering accounts for nearly 50% of agentic activity." The *directional* idea (early agentic deployments are concentrated in coding, DevOps, and data workflows) matches general commentary, but the precise "nearly 50%" figure should be treated as a **single-article statistic, not industry consensus.** Downstream agents should attribute it to the article rather than presenting it as an established benchmark.