New research from RWS (RWS.L) has exposed a critical fault line in enterprise AI deployment: the vast majority of content leaders have little to no confidence that AI can handle cultural and emotional nuance across global markets – yet they are scaling AI-generated content across those same markets regardless. The result is a deepening financial and operational crisis that speed, volume and model upgrades alone cannot fix.
The study, based on a survey of 200 senior content leaders across the US, UK and Asia-Pacific, finds that while 86% say AI has accelerated content creation, 65% report it has simultaneously slowed localization – generating a compounding rework burden that is consuming over a fifth of enterprise localization budgets every year.
A growing financial burden
Over a fifth (21%) of localization budgets are lost to rework: content that must be corrected or culturally adapted before it can be deployed in global markets. For a global enterprise spending $5 million on localization, that represents over $1 million lost annually fixing AI’s own outputs from generic models.
“Ask content leaders whether AI can truly handle cultural nuance, and fewer than one in ten say yes with confidence. Yet the same leaders are scaling AI-generated content across global markets regardless,” said Emma Fisher, VP of Global Marketing at RWS. “The answer isn’t to slow down – it’s to deploy smarter AI. AI that understands context, culture and brand intent as fluently as it generates content at scale.”
The Cultural Intelligence gap
The research exposes a striking disconnect in how enterprises are deploying AI. 71% of content leaders are using generative AI for translation – a task AI handles with increasing competence. Only 20% use it for localization, which demands something far harder: the cultural fluency to make content feel native, resonant and relevant in every market. Implementing AI at scale into these already-strained workflows does not resolve the problem. It accelerates it.
A dangerous complacency
Perhaps most striking is the confidence gap the research reveals. Despite 56% of leaders describing their organizations as “managing but stretched,” and only one in six saying they are handling content demands well, more than half believe they will cope better in three years’ time – without significant structural change. RWS’s research suggests this optimism is misplaced. As AI-generated content volumes continue to grow across more languages, formats and channels, the complexity tax will only compound – unless enterprises embed cultural intelligence into their content operations upstream, before the damage is done.
The research is based on a survey of 200 senior content leaders across the US, UK and Asia-Pacific. Respondents represent large global enterprises responsible for content strategy, localization and digital communications.
