---
title: "Agentic development: software vendors advocate a multiplayer approach — Marsanne Studio AI watch"
description: "Software vendors (Upsun, Slack, Atlassian, GitLab, GitHub) are organising agentic software development around human-agent collaboration, despite persistent tool fragmentation."
url: "https://marsanne.studio/en/ai-watch/agentic-development-software-vendors-advocate-a-multiplayer-0e78a1/"
lang: "en"
---

[Marsanne Studio](https://marsanne.studio/en/)

[← Back to home](https://marsanne.studio/en/ai-watch)

[← The AI watch](https://marsanne.studio/en/ai-watch)

Analysis · 7 October 2026

# Agentic development: software vendors advocate a multiplayer approach

Original title, in French: “Développement agentique : les éditeurs prônent le multijoueur”

**Publication**

LeMagIT (France) · subscribers only

**Author**

Gaétan Raoul

**Published on**

7 October 2026 · read in the page structured data

## Summary

Software vendors (Upsun, Slack, Atlassian, GitLab, GitHub, Anthropic) are organising agentic development around a "multiplayer" model in which humans and AI agents collaborate, rather than around solo practice. At Upsun, the engineering and product teams have merged; at ServiceNow, three previously separate teams are now orchestrated together. Slack Code, generally available since August 2026, allows teams to create code channels where agents (Claude Code, Devin, GitHub Copilot, Vercel) operate under human supervision; Anthropic states that 60 to 65% of its pull requests are approved via Slack. Atlassian has launched the "Agentic Multiplayer Protocol" to integrate agents into Jira, Confluence and Loom. A GitLab study of 1,528 professionals shows that only 28% have fully integrated tools, and that 83% of DevSecOps profiles regard the build-up of generated code as a risk. StackOverflow's 2026 Developer Survey highlights a persistent problem with attributing the sources of code for 93% of the respondents concerned.

Summary based on the full text of the article.

[Read the article on LeMagIT →](https://www.lemagit.fr/actualites/366652015/Developpement-agentique-les-editeurs-pronent-le-multijoueur)

## Figures cited

-   62.5 %
    
    Share of Anthropic's pull requests approved via Slack
    
    Anthropic, internal pull requests, 2026
    
    > « Anthropic affirme que 60 à 65 % de leurs pull requests sont validées via Slack »
    
    Source: Anthropic
    
-   28 %
    
    Share of professionals with fully integrated software development tools
    
    1,528 professionals surveyed, GitLab study, 2026
    
    > « GitLab note que seuls 28 % des sondés ont des outils de développement logiciel entièrement intégrés, alimentés par des données et des flux de travail partagés »
    
    Source: GitLab
    
-   60 %
    
    Share of respondents who believe agentic development makes it possible to exceed expectations
    
    1,528 professionals surveyed, GitLab study, 2026
    
    > « le développement agentique permet de dépasser les attentes (60 %) »
    
    Source: GitLab
    
-   78 %
    
    Share of respondents who believe agentic development speeds up code approval
    
    1,528 professionals surveyed, GitLab study, 2026
    
    > « d’accélérer la validation de code (78 %) »
    
    Source: GitLab
    
-   83 %
    
    Share of DevSecOps profiles who see the build-up of generated code as a risk
    
    1,528 professionals surveyed, GitLab study, 2026, DevSecOps profiles
    
    > « la majorité des profils DevSecOps (83 %) voit l’accumulation de code généré comme un risque »
    
    Source: GitLab
    
-   93 %
    
    Share of developers for whom the problem of attributing code sources remains significant
    
    13,945 developers surveyed out of more than 30,000, Developer Survey 2026, StackOverflow
    
    > « ce problème d’attribution des sources demeure important pour 93 % des 13 945 personnes sondées »
    
    Source: StackOverflow Developer Survey 2026
    
-   87 %
    
    Share of developers who trust AI-generated results for verifiable, low-criticality tasks
    
    Developer Survey 2026, StackOverflow, more than 30,000 developers surveyed
    
    > « 87 % des sondés font confiance aux résultats générés, mais principalement pour les tâches vérifiables et peu critiques »
    
    Source: StackOverflow Developer Survey 2026
    

## Classification

**Country**

Worldwide scope

**Company size**

All sizes

**Industries**

Software vendors, IT services companies, telecoms

**Companies**

ServiceNow (deploys AI), Slack (supplier), CodeRabbit (supplier), Salesforce (cited), Atlassian (supplier), GitHub (supplier), Microsoft Research (cited), GitLab (study author), Lovable (supplier), Vercel (supplier), Mistral (supplier), Replit (supplier), Anthropic (supplier), Snowflake (cited), StackOverflow (study author), Upsun (supplier), SensioLabs (supplier)

**Keywords**

agentic development, AI agents, code review, Slack Code, software development, agent orchestration

Entry written by Marsanne Studio from the cited article, with the help of an AI model (claude-sonnet-5). The rights to the article and its illustrations belong to LeMagIT and its author. Entry translated from French by Claude Sonnet 5.5.