Workplace Language Access Without Compromising Trust

Every organisation says its people are its greatest asset. Far fewer can say that all of those people hear the same message, in a language they fully understand, at the same time, and feel equally able to respond.

For decades, professional interpretation has followed the budget. It has been concentrated where spending and visibility are greatest: the shareholder meeting, the board session, the flagship conference. Everyday workplace communications, such as town halls, training, safety briefings, company updates and regional meetings, have largely gone without. Not because they matter less, but because interpreting every one of them has rarely been financially practical.

In this article, we look at how big that gap really is, how AI interpretation changes the economics of workplace language access, and which safeguards HR leaders need so that wider access does not come at the expense of security, reliability or trust.

How Big is the Workplace Language Gap?

The gap is bigger than it looks. According to the International Labour Organization, international migrants made up 4.7% of the global labour force in 2022, some 167.7 million workers. Add the many teams that work across borders every day, and a large share of the workforce routinely receives important information in a second or third language, if it reaches them at all.

In the UK, the 2021 Census for England and Wales found that 5.1 million people, 8.9% of usual residents aged three and over, had a main language other than English or Welsh. Most spoke English well or very well, but more than a million said they could not speak it well or at all. And speaking a language well in daily life is not the same as following a restructuring announcement, a pension change or a new safety procedure in it, live, with no chance to pause and reread.

The Cost of the Gap for Multilingual Employees

When language support is rationed, multilingual employees pay the price. They follow a strategy update through a colleague's summary, sit through compliance training they only partly understand, or stay silent in the Q&A because they can't ask a question in their own language. Over time, that affects engagement, safety and retention, and it quietly undermines the inclusion commitments many organisations make in public. It also risks creating a two-tier workforce, where fluency in the corporate language decides who is fully informed.

The gap is often widest among frontline and shift workers in sectors such as logistics, manufacturing, hospitality and healthcare, where multilingual teams are common and missed information has immediate consequences. These employees are also the least likely to have the time or the channels to catch up later.

For HR leaders, this is not a niche accessibility issue. It is a question of whether every employee has equivalent access to the information that shapes their work.

How AI Interpretation Changes the Economics

AI interpretation changes that equation. Where a human-interpreted session involves booking, scheduling and per-language costs, AI can deliver live speech translation across many languages at a fraction of the cost, often at short notice. That makes it realistic to interpret not just the annual all-hands, but the weekly team update, the onboarding session, the regional briefing, and everyday external communication.

In practice, AI handles volume. It lets organisations extend language support to communications that previously had none, and to languages that were never budgeted for. It also removes practical barriers beyond cost. Some languages have few available interpreters, and short-notice meetings rarely leave time to book them, whereas AI can be ready whenever a session is scheduled. Nor does it need special equipment: at the end of 2025, 44% of audiences on the Interprefy platform listened through a mobile app and 28% through a web browser, while only 6% used traditional on-site receivers.

Oddmund Braaten, CEO of Interprefy, often describes interpretation as a pyramid:

At the top, professional interpreters cover perhaps 10% of an organisation's need: the high-level meetings at headquarters. The other 90% happens further down and further out, in regional offices and day-to-day meetings where budgets for interpreters rarely exist. That is where AI changes the picture.

Interprefy's own figures reflect that shift. In the second quarter of 2026, AI was involved in 69% of the language services sessions we delivered, and AI-only sessions grew by almost a third year on year. Much of that growth came from settings that previously had no language support at all.

Democratisation is Not a Race to the Bottom

However, wider access only counts as progress if standards hold. The risk in any market that suddenly becomes cheaper is that buyers compare on price alone and vendors cut corners to compete. In interpretation, those corners are security, reliability and quality.

Workplace communications are often sensitive: restructuring announcements, financial results, policy changes. An AI tool that is cheap but vague about where audio is processed, how data is stored or who can access recordings isn't democratising anything. Instead, it is shifting risk onto the employer and its employees. Recordings and transcripts of staff meetings are personal data, and the employer remains accountable for how they are handled under UK GDPR.

The questions to ask are the ones IT and security teams already ask of any supplier. At Interprefy, audio is protected with 256-bit encryption and data is stored in Europe, and we regularly work through compliance reviews with the IT departments of pharmaceutical companies and other regulated organisations. Every provider in this market should be able to answer those questions just as clearly.

Reliability matters just as much. A town hall that drops out halfway through, or a translation that garbles a policy change, damages trust more than no interpretation at all. Employees who receive a confusing message in their own language are unlikely to rely on it next time.

So the bar for AI should be the one organisations already apply to established interpretation services: enterprise-grade security and data protection, clear policies on data use, proven platform stability, technical support during live events and transparency about how quality is measured.

Where to Draw the Line Between AI and Professional Interpreters

Democratisation also means being honest about what AI is suited to. It performs best where speech is structured and the setting predictable: presentations, updates, training. Professional interpreters remain the right choice where nuance, cultural sensitivity or legal and personal consequences are at stake, such as works council negotiations, grievance processes or conversations about health and wellbeing.

That decision should be made session by session and language by language, based on content, format and audience, rather than defaulting to whichever option is cheapest. Getting that split right is part of the value of AI interpretation, not a limitation of it.

Safeguards that build employee trust

Trust is ultimately what employees experience, not what a vendor promises. HR leaders can protect it in a few practical ways:

• Hold AI interpretation providers to the same security and compliance requirements as the rest of the organisation's technology, including where data is processed and stored.
• Prepare sessions properly: share agendas, speaker names, acronyms and terminology in advance, including terms that should not be translated.
• Test audio and set-up before live events, with technical support on hand.
• Tell employees when AI interpretation is being used, so they can judge the information accordingly and raise concerns.
• Ask for feedback in employees' own languages, and act on it.

Finally, encourage employees to treat AI interpretation like any other AI tool. It will not be perfect, and neither are human interpreters. It is there to help people understand, and it works best when they still apply their own judgement.

Bigger than cost savings

The real opportunity is not simply cheaper interpretation. It is the chance to make language access a standard part of workplace communication, rather than a privilege reserved for a handful of high-profile moments. Done responsibly, AI can help organisations reach every employee and partner in a language they understand. Done carelessly, it risks eroding the very trust it was meant to build.

Interprefy's mission has always been to democratise interpretation: to make it available and affordable, while keeping it secure and professional. The organisations that get this right will widen access and hold their standards at the same time. That combination, not cost alone, is what real progress in language access looks like.

Sources

• International Labour Organization (2024). Global Estimates on International Migrant Workers, published 18 December 2024. Full report (PDF).
• Office for National Statistics (2022). Language, England and Wales: Census 2021, released 29 November 2022.