Analysis ·
AI in business: unclear responsibilities are holding back adoption
Original title, in French: “IA en entreprise : le flou des responsabilités bloque l’adoption”
- Publication
- Brief IA (France)
- Author
- Tom Levy
- Published on
- 1 October 2026 · read in the page structured data
Summary
According to an analysis by PwC published on 1 October 2026, companies are struggling to decide who is responsible for AI systems among the CIO, CTO, CISO and AI departments. This lack of clarity is slowing scaling, while McKinsey and Stanford HAI note that 88% of companies used AI in 2025 but only 39% saw a measurable impact. Deloitte reports that 80% of automation leaders want to accelerate investment in AI agents, while only 21% of organisations have mature governance. A survey by OneTrust and Sapio Research shows that only 5% of organisations have clear responsibility across the whole AI lifecycle. Okta and IBM recommend treating agents as users with limited identity and access.
Summary based on the full text of the article.
Read the article on Brief IA →Figures cited
-
88 %
share of companies using AI in at least one function in 2025
companies, 2025, data relayed by McKinsey and Stanford HAI
« En 2025, 88 % des entreprises utilisaient déjà l’IA dans au moins une fonction, mais seules 39 % déclaraient un impact mesurable sur leur résultat opérationnel »
Source: McKinsey et Stanford HAI
-
39 %
share of companies reporting a measurable impact on operating profit
companies, 2025
« seules 39 % déclaraient un impact mesurable sur leur résultat opérationnel »
Source: McKinsey et Stanford HAI
-
80 %
automation leaders planning to accelerate investment in AI agents
automation leaders
« Deloitte indique que 80 % des responsables de l’automatisation prévoient d’accélérer leurs investissements dans les agents d’IA. »
Source: Deloitte
-
21 %
organisations reporting mature agentic governance
organisations surveyed by Deloitte
« seuls 21 % des organisations interrogées déclarent disposer de capacités de gouvernance agentique suffisamment matures »
Source: Deloitte
-
74 %
organisations that have reached departmental adoption or scaled AI
1,200 decision-makers in eight markets, OneTrust and Sapio Research survey
« 74 % des organisations interrogées déclarent avoir atteint un niveau d’adoption départemental ou une mise à l’échelle »
Source: OneTrust et Sapio Research
-
52 %
organisations using AI in several functions or embedded in operations
1,200 decision-makers in eight markets
« 52 % utilisent l’IA dans plusieurs fonctions ou l’ont intégrée à leurs opérations »
Source: OneTrust et Sapio Research
-
5 %
organisations with clear coordination and accountability across the AI lifecycle
1,200 decision-makers in eight markets
« seules 5 % indiquent que la coordination et la responsabilité sont claires sur l’ensemble du cycle de vie de l’IA »
Source: OneTrust et Sapio Research
-
— %
advanced companies committing to a common enterprise AI platform
most advanced companies
« plus des deux tiers des entreprises considérées comme les plus avancées s’engagent vers une plateforme IA d’entreprise reposant sur une architecture commune et un plan de contrôle partagé »
Source: BCG
Classification
- Country
- Worldwide scope
- Company size
- All sizes
- Industries
- All industries, or industry not specified
- Companies
- Okta (cited), McKinsey (cited), Deloitte (study author), PwC (study author), Sapio Research (study author), Stanford HAI (cited), BCG (study author), IBM (cited), OneTrust (study author)
- Keywords
- ai governance, ai agents, accountability, security, ai adoption, shadow ai
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 Brief IA and its author. Entry translated from French by Claude Sonnet 5.5.