Business & Technology
RWS study finds multilingual AI gaps are narrowing
RWS has published results from a TrainAI study on multilingual synthetic data generation by large language models. The research found that performance gaps between major and underrepresented languages have narrowed.
Results can also shift markedly between model releases, with changes in quality and cost-related measures that may alter which system is best suited to a given task.
The findings examine how leading large language models perform when generating synthetic data across multiple languages, an area drawing increased attention as companies deploy AI tools in global content and customer-facing workflows.
One of the clearest trends is improved output in languages that have historically received less support from mainstream models. According to the research, the gap between English and lower-resource languages has narrowed significantly in the latest generation of systems tested.
Among the examples cited, Google’s Gemini Pro recorded quality scores above 4.5 out of 5 in Kinyarwanda, a language in which earlier model versions had struggled to generate coherent text. The study also pointed to broader gains across the market, including improvements in GPT and Claude Sonnet models.
Benchmark Drift
Alongside those gains, researchers highlighted what they called benchmark drift. In practice, that means a newer version of a model can perform worse than an earlier one on some tasks, even as the broader market trend points to improvement.
The latest version of GPT trailed smaller models on several content generation tasks where its predecessor had remained competitive. The study also found that tokenizer efficiency, which affects how much text can be processed for a given cost, varied sharply between successive releases.
That matters for businesses comparing systems for translation, content creation and other multilingual applications, because a model upgrade may change both output quality and economics. A model family that performs well in one release may lose ground in the next, while a rival system may improve in specific languages or workloads.
Public rankings alone are not enough for model selection, particularly for companies operating across multiple markets and language groups. Instead, the research recommends testing models repeatedly against specific use cases as new versions become available.
Enterprise Use
The study’s conclusions are aimed at companies building AI into production workflows rather than conducting one-off experiments. For those users, differences in cultural nuance, text coherence and token efficiency can shape both customer experience and operating cost.
RWS, which provides language, content and AI-related services, framed the findings as a case for closer oversight of model choice and for retaining human review in multilingual deployments. It said high-quality, culturally nuanced data remains central to assessing whether a model is suitable for a specific market or business function.
Vasagi Kothandapani, Chief Executive Officer of TrainAI by RWS, set out that position in a statement accompanying the findings. “This study signals a transformative moment that’s not about replacing human expertise, but about elevating it with the right technology,” Kothandapani said.
“As AI becomes more capable across languages, the need for deep cultural intelligence and human validation is more critical than ever. This is why RWS is guiding enterprises into this new reality by integrating these powerful technologies into content workflows with experts in the loop to ensure accuracy, cultural resonance, and brand consistency on a global scale.”
The research also examined less visible technical measures that can influence adoption decisions. One is tokenizer efficiency, which affects the number of tokens a model uses to represent text and therefore has a direct bearing on usage costs.
Tomáš Burkert, Head of Innovation at TrainAI by RWS, said that issue is often overlooked when businesses compare systems. “A model’s real-world value often comes down to specific, frequently overlooked metrics,” Burkert said.
“Factors like tokenizer efficiency, which can make one model 3.5 times more cost-effective than another in certain languages, are critical. The foundation of a successful AI strategy is a continuous validation process, rooted in high-quality, culturally nuanced AI data, to ensure you’re not just adopting a model, but the optimal model to address your unique enterprise requirements.”
RWS said the overall direction of travel in multilingual AI remains positive, with stronger performance across a wider range of languages than in earlier generations. But the central argument is that improvement is uneven and that model selection should be treated as an ongoing process rather than a one-time decision.
The report recommends independent evaluation with each new release to check whether a model still matches the intended use case, particularly when language coverage, quality thresholds and cost control are all under review.
Business & Technology
Thames Water leakage targets are ‘not realistic’ says boss
Chris Weston, speaking on the BBC’s Big Boss Interview podcast, stated that some of the firm’s targets were beyond what they could achieve.
He said: “We have to hit a certain level of leakage, but it is so far in excess of what we are capable of doing, I think anyone would be capable of doing, however much money you invested, that it is not going to be achievable.”
Thames Water, the largest water company in the UK, has been under fire recently for its handling of sewage discharges and leaks.
Last year, it was fined a record £122.7 million by regulator Ofwat, largely for breaching sewage spill rules.
The company, serving 16 million customers in London and parts of southern England, treats 4.3 billion litres of waste daily.
Mr Weston mentioned that “99.5% of the time” the waste is treated successfully, although “sometimes something goes wrong”.
He added that while the company wants to improve on pollution, the chance of getting to zero pollution was “very, very slim”.
Ofwat responded to Mr Weston’s comments by stating: “With around a fifth of water put into supply still lost through leakage, companies must deliver on the commitments they have been funded to achieve.
READ MORE: Real reason for David and Victoria Beckham’s fallout with Brooklyn revealed
(Image: PA)
“Water company targets are intended to be ambitious and drive better outcomes for customers and the environment.”
The Environment Agency spokesperson said the agency and the public expect Thames Water to comply with the law, adding that the agency would continue to hold companies to account where performance falls short.
James Wallace, chief executive of campaign group River Action, criticised Thames Water.
He said: “Thames Water’s tactics of opacity and deflection fool no-one.
“Telling the public to save water while this wasteful profit-obsessed corporation leaks 570 million litres of treated drinking water every day is offensive.
“There is nothing ‘realistic’ about accepting sewage pollution as inevitable.”
Thames Water has been grappling with billions of pounds in debt and faces the risk of temporary nationalisation, known as a special administration regime.
Mr Weston, however, warned of the potential burden on taxpayers if this were to happen.
Instead, he backed a rescue deal proposed by the firm’s lenders.
Meanwhile, Mr Weston defended the company’s pay levels, as his pay rose by 14% to £1.163 million in the year to March, while other directors received bonuses totalling £4.1 million.
Business & Technology
telent urges UK defence to prioritise tech integration
Telent has urged the UK defence sector to place greater emphasis on integrating new technologies with existing systems, an argument set out by Strategy and Development Director Barry Zielinski.
Investment in artificial intelligence, autonomous systems and other emerging technologies will not deliver a meaningful operational advantage unless those tools can work together within a wider operational framework, Zielinski said.
He argued that the operating environment is shifting quickly as adversaries adapt faster, technology cycles shorten and the line between physical and digital battlefields becomes less distinct. In that context, military success depends less on the strength of any single platform or sensor and more on the speed at which armed forces can connect information, decisions and operational effects across domains.
The comments reflect a wider debate in defence procurement and military planning, as governments and suppliers pay more attention to artificial intelligence, cyber tools, advanced sensors and space-based assets. In Zielinski’s view, these systems should be judged not in isolation but by how well they connect with networks, data systems, command structures and personnel.
Integration focus
Communications networks, operational facilities and digital systems form the foundation of modern military capability because they allow information to move securely and reliably, according to Zielinski. He said resilience and security must be built in from the start through approaches such as secure-by-design and zero-trust principles.
That view places infrastructure at the centre of defence modernisation rather than treating it as a support function. It also shifts part of the discussion away from procuring new tools and towards ensuring that existing and new assets can share data and support decisions coherently.
Artificial intelligence has become a major topic in defence because of its potential to process information and support decision-making. Zielinski said that potential depends on more basic conditions, including data quality and the resilience of the infrastructure that carries it.
If communications, trusted data and secure networks are unavailable or compromised, the value of more advanced systems is reduced, he said. In practice, that means creating decision advantage is not only about software or platforms, but also about the systems and physical assets that underpin them.
Human role
Zielinski also addressed automation, which is drawing attention across defence organisations seeking greater efficiency and a faster operational tempo. He said its main value often lies in reducing repetitive work rather than replacing people.
He cited predictive maintenance, autonomous monitoring, automated network management and logistics optimisation as examples. Those uses can free skilled personnel to focus on tasks where judgement and experience remain essential, he said.
“The most effective technologies do not replace human capability – they amplify it,” Zielinski said.
The argument comes as defence planners increasingly talk about integrated operations across land, sea, air, cyber and space. In that model, infrastructure such as communications, transport, energy and digital systems becomes more strategically important because it links the movement of people, information and resources.
Zielinski pointed to the Falkland Islands as an example of how infrastructure supports long-term readiness, citing runway infrastructure as part of maintaining strategic capability. The example illustrated how closely linked infrastructure and operational output are becoming.
From pilots to deployment
Zielinski also argued that the UK already has access to much of the technology needed for future operations. The central question, he said, is whether those technologies can be integrated and adopted quickly enough to meet operational needs.
That places emphasis on collaboration between government, industry, academia, small and medium-sized enterprises and the Armed Forces. Combining those perspectives can improve the practical design of technology and help integrate it into future capability, he said.
He set out three priorities for defence organisations: connecting networks, data, platforms, people and infrastructure; improving collaboration across the sector; and moving beyond demonstrations, pilot schemes and concepts so new systems reach operators more quickly.
The broader message is that military advantage is likely to depend on how effectively defence organisations bring together people, infrastructure, data and digital systems, rather than on who has the best single piece of equipment. “The question is not whether the technology is available. The question is whether it can be integrated quickly enough to provide the Armed Forces with a genuine operational advantage,” Zielinski said.
Business & Technology
Alphatax expands transfer pricing software with two buys
SOFIAH NICHOLE SALIVIO
News Editor
Alphatax has acquired TP Accurate and Intra Pricing Solutions, expanding its transfer pricing software offering.
The acquisitions add two products to the group’s tax technology portfolio. TP Accurate develops software for intra-group financing arrangements, while Intra Pricing Solutions offers TPGenie, a tool that uses automation and artificial intelligence to produce transfer pricing documentation for multinational companies and advisers.
Transfer pricing has become a growing focus for tax departments as regulators increase scrutiny of how multinational groups price transactions between related entities. Companies often manage financing analysis, documentation and compliance through separate systems, leaving tax teams with fragmented processes and greater audit exposure.
The additions broaden Alphatax’s coverage across the transfer pricing workflow. The products are intended to help customers manage compliance more efficiently, strengthen governance and reduce audit risk.
The transactions also form part of a broader platform strategy at the business, previously known as Tax Systems. Alphatax’s long-term goal is to build a single operating system for tax, bringing together compliance areas that have traditionally been managed through standalone tools.
Bruce Martin, Chief Executive Officer of Alphatax, said the deals supported that strategy.
“This is another important step in our growth strategy and reflects our continued investment in creating a more connected future for tax,” said Bruce Martin, Chief Executive Officer, Alphatax.
“Transfer pricing is one of the most complex areas of tax compliance, with increasing regulatory scrutiny and growing demands on tax teams. Bringing TP Accurate and Intra Pricing Solutions into Alphatax adds market-leading capabilities that deliver immediate value for customers while accelerating our vision of the world’s first tax operating system. We’re delighted to welcome both teams to Alphatax,” Martin added.
Product fit
The rationale for the two acquisitions lies in different parts of the transfer pricing process. One addresses financial transactions within corporate groups, including loans, guarantees and other financing arrangements that require detailed pricing analysis. The other focuses on preparing the documentation companies need to support their transfer pricing positions.
For large multinationals, those tasks have become more burdensome as tax authorities demand more detailed support for cross-border arrangements. Software providers have responded by building tools that automate calculations, standardise reporting and help companies maintain records across jurisdictions.
Intra Pricing Solutions’ management said the transaction would allow it to continue developing its software within a larger platform.
“We were looking for a partner that shares our long-term vision for innovation in transfer pricing,” said Arjen Rommens, Co-Founder & CTO, Intra Pricing Solutions.
“In Alphatax, we found exactly that. Together we can invest more, innovate faster and continue supporting our customers for many years to come,” Rommens added.
TP Accurate was founded to address pricing for intercompany financial transactions, a niche but significant part of the market. Such transactions can be difficult to assess because companies must justify the terms applied between related parties as though they had been agreed by independent entities.
Michael Vorndran, Founder of TP Accurate, said that focus would now sit within a larger organisation with a broader customer base.
“I founded TP Accurate to address a longstanding gap in transfer pricing technology: the ability to accurately price intercompany financial transactions, which run into the trillions of dollars annually,” said Michael Vorndran, Founder, TP Accurate.
“Joining Alphatax means many more companies around the world can benefit from our solution and we’re proud to be part of a team like Alphatax,” Vorndran added.
Market position
Alphatax is backed by Providence Equity Partners and sells tax and accounting software to large companies and advisory firms. It says it works with more than 42% of the FTSE 100 and 80% of the top advisory firms, while more than 30,000 tax professionals have been trained to use its software.
The group has operated for more than three decades and has sought to widen its reach across tax compliance functions as companies digitise processes once handled through spreadsheets and localised applications. More than 200,000 submissions are filed each year using its systems, according to the company.
By acquiring specialist providers rather than building every function in-house, Alphatax is following a path taken by many software groups seeking to assemble broader platforms in niche business markets. In tax technology, where rules differ by jurisdiction and compliance demands shift regularly, vendors argue that integrated tools can help companies manage risk more consistently across their operations.
The two acquisitions place greater emphasis on transfer pricing, an area that remains both technically complex and commercially important for multinational groups with cross-border financing and intercompany trading arrangements.
-
Business & Technology3 weeks agoHSBC UK & Visa test AI shopping with live payments
-
Business & Technology3 weeks agoValarian lands USD $50 million backing for sovereign AI
-
Business & Technology4 weeks agoMouser warns against viral hacks to cool overheating phones
-
Oxford News4 weeks agoNew romantasy bookshop attracts queues of customers
-
Business & Technology4 weeks agoSNP & Palantir launch AI tools for SAP transformations
-
Business & Technology4 weeks agoKane tops England influencer rankings after Mexico win
-
Traffic & Transport4 weeks ago‘I felt my spine and body split’: the woman who was hit by a child on a Lime bike – and denied compensation | Ebikes
-
Oxford News4 weeks agoDWP now checking bank accounts for Universal Credit and Pension Credit
