Are Enterprise Translation Management Systems facing a BlackBerry moment?

In this article, we reveal the extraordinary impact AI is having on traditional translation business models, asking if legacy translation tools are about to experience a Blackberry moment and become obsolete.

"BlackBerry moment" refers to a turning point in the tech industry where a dominant market leader fails to adapt to a new, disruptive technology, resulting in a rapid loss of market share.

Enterprise Translation Management Systems (TMS) are undergoing the most profound structural change in decades. What was once a stable market built on workflow software, and licensing models is now being reshaped by agentic AI, requirements for sovereign AI deployments, and demand for secure human-in-the-loop solutions to mitigate risk by verifying AI outputs. In short, global customers don’t want clunky, expensive translation management, they are seeking self-service AI translation and verification at scale.

This shift is no longer theoretical. Market data, investor reactions, and industry consolidation all draw the same conclusion: the traditional TMS model is under pressure, as customers increasingly seek proprietary platforms built for the AI Age, not retrofitted products.

Market evidence: Legacy translation models

Large language service providers (LSPs) that once invested heavily into TMS tools now appear to be under pressure to urgently pivot to a new model of working. This means developing or buying-in new AI solutions that give customers the convenience and cost savings of self-service AI translation, plus the benefits of sovereign deployment with automated human verification when needed.

RWS, one of the world’s largest language technology providers and the owner of Language Weaver, the enterprise TMS provides a useful barometer for the broader market.

In FY2025, RWS reported declining revenues, sharply reduced profits, a significant reported loss driven by goodwill impairments, and a material dividend cut to fund an accelerated AI-first strategy.

Over the same period, its share price fell more than 40% from prior highs, reflecting investor concern around margin pressure and disruption to traditional licensing revenues.

RWS leadership has been explicit that this is a structural transition rather than a cyclical downturn, reorganising the business around AI-led segments and rapidly developing new tools to adapt to changing buyer behaviour.

The TMS licensing model is being rewritten

Traditional TMS platforms historically benefited from predictable enterprise licences, long implementation cycles, and deeply embedded workflows. Generative AI challenges all three.

As AI enables on-demand, self-service translation at scale, value shifts away from workflow management and toward governance, assurance, and measurable business outcomes. Legacy providers are now racing to rebalance their revenue mix toward SaaS and AI-driven solutions.

This urgency to develop and introduce the new secure AI products that customers require is inspiring some large LSPs to acquire solutions as an alternative to developing in-house.

Industry signal: Unbabel acquired by TransPerfect

In August 2025, US giant TransPerfect acquired Unbabel, an AI-native translation platform developed in Portugal. The acquisition brought translation-specific Large Language Models and industry-standard quality evaluation tools into TransPerfect’s existing technology stack, underlining the strategic importance of AI-native platforms.

This consolidation illustrates how customers now expect self-service AI translation options. It also shows how the benefits of buying in an already developed product suite rather than building one from scratch that may not work is influencing buying decisions as AI adoption accelerates.

But do Enterprise buyers really need legacy TMS with new add-ons at all when popular AI tools are advancing so quickly?

Self-service disruption: ChatGPT & CoPilot

The sharp fall in share prices of traditional translation companies has aligned with the introduction of tools like ChatGPT and Copilot. Investors quickly understood how these new tools can disrupt legacy systems because they generate instant multilingual output, remove onboarding friction associated with buying a TMS, and free users from the restrictions of costly licensing models.

At the same time, investors in traditional businesses know that pivoting to developing new AI solutions is costly, takes time, carries a high risk of failure, and can mean writing off old legacy software that may be securitized, resulting in losses.

However, ChatGPT and Copilot in their current form create problems for enterprise users. There are well-documented risks around AI making things up to please you (hallucinations), bias, lack of auditability, and challenges verifying the outputs. Thus, such tools are unsuitable for regulated or brand-critical content without ironclad governance controls such as human-in-the-loop.

In short, buyers want sovereign AI that provide accurate, domain-specific translations and allows experts to efficiently verify results within the workflow.

Sovereign AI with verification: GAI Translate

Now, AI development companies like Guildhawk are entering the market to meet the increasing demand for AI solutions that meet requirements for security, interoperability and human verification.

For example, from the outset, GAI Translate was developed as an AI translation solution for high-risk settings, where accuracy and human verification are essential. Rather than retrofitting AI into legacy infrastructure, it is built natively around sovereign deployment, agentic orchestration, automated expert human assurance, and auditability.

Consequently, low-risk content flows at AI speed. High-risk AI output is automatically escalated to a vetted human expert for verification or certification. The result is outputs that are defensible, brand-safe, and compliant.

GAI Translate is unique because it is a British proprietary solution developed in partnership with users to solve their specific problems. Deployment is quick, scalable and low risk because the products are fully developed, tested and trusted by professionals in high-stakes industries.

Creating ethical, sustainable AI solutions that deliver the results partners require calls for continuous investment into research and development, the kind of pioneering work conducted by Sheffield Hallam University and Guildhawk.

Conclusion: AI safe by design

This is not a collapse of the traditional translation business models, but a reset of expectations. Platforms architected for intelligent autonomy and trust will define the next standard. GAI represents this next chapter: AI designed for enterprise growth, interoperability, and human verification of outputs.