Monmouthshire and Torfaen could use AI Welsh translations
Councils, and other public bodies, are legally required to produce documents and provide services in Welsh.
Monmouthshire County Council looked at alternatives to its current use of external translators as part of its budget setting process due to increased demand which has seen spending on translations exceed the allocated budget.
Its performance and overview scrutiny committee was told total costs, to the council of providing Welsh language services, was £226,940 last year.
But it wasn't clear if that figure only covered translations or also the cost of two officers after equalities and Welsh language manager Pennie Walker described the 'major costs' as translations and her salary and that of the Welsh language officer while ensuring compliance with Welsh language standards is the 'day to day responsibility of all officers'.
Nia Roberts, the Welsh language officer, said closer working with neighbouring Torfaen Borough Council is currently 'the most desired option' to save on translation costs and it is already using AI, also known as artificial intelligence.
Ms Roberts said Monmouthshire currently uses external translators, who may already be using AI systems, but that wouldn't produce a saving for the council as it currently pays by the word.
She said AI would 'never produce 100 per cent accurate translations' and added: 'It will need to have some kind of proof reading to make sure the translation is accurate.'
Rogiet Labour member Peter Strong had asked if a favoured option had emerged from the ongoing review that has also considered continuing with external translators and setting up its own in-house translation team.
Ms Roberts said joining with Torfaen 'looks the more desirable' and said: 'Torfaen is similar to Monmouthshire in the type of documents to be translated.'
She also said Torfaen uses technology that memorises words it has previously translated which would help with the consistency of documents.
Torfaen councillors, who were presented with their annual report on how their council is complying with Welsh language standards at their June meeting, were told AI, has helped with increasing translations by 24 per cent on the previous year.
Torfaen's Welsh language officer Alan Vernon-Jones cautioned: 'Everything needs to be proof read by a competent Welsh speaker.'
Monmouthshire's review of translations has also given 'careful consideration' to the potential impact on small Welsh businesses, quality and timeliness, and the need to maintain full compliance with the Welsh language standards.
Conservative councillor for Gobion Fawr Alistair Neill asked Ms Roberts if consideration had been given to working with other public services such as the police and NHS.
The officer said Welsh language officers in public services across South East Wales have a working group to share best practice which was 'more broad than just translations.'
On AI Cllr Neill, who chairs the committee, said: 'It sounds like AI doesn't speak Welsh very well. The 'tech bros' are going to have to do some more work on it.'
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Finextra
2 hours ago
- Finextra
Why the Smartest Fintechs Are Scaling with AI Agents – Not Headcount: By David Weinstein
For the better part of a decade, fintech growth has followed a familiar trajectory: secure funding, hire aggressively, and scale fast in pursuit of market traction. It worked. High-performing teams, ambitious roadmaps, and well-capitalised burn rates became the standard operating model for any startup with global aspirations. But that playbook is starting to look outdated. Today's most forward-thinking fintechs are flipping the script. Instead of scaling with people or piecemeal software, today's most advanced fintechs are scaling with context-aware AI infrastructure, enabling autonomous agents to operate with memory, relevance, and the ability to adapt across time. In other words, the smartest fintechs aren't just hiring more people, they're designing for a world of leverage. From Chatbots to Autonomous Operators To be clear, this isn't about adding another chatbot to the support queue or slapping GPT on top of a FAQ. The new generation of AI agents are far more capable. These aren't just reactive tools dropped into workflows - they're embedded, active participants in how work gets done. They're not replacing human judgment, but taking over the repetitive execution that bogs it down. By operating within a structured, evolving knowledge graph, these agents access the right context, perform tasks across systems, and maintain continuity over time so that human operators can stay focused on what matters: discernment, creativity, and strategic direction. Imagine an agent that scans customer interactions across CRM, support, and marketing tools, then identifies churn risks and recommends retention strategies - autonomously. Or a compliance agent that tracks regulatory changes, audits internal data for alignment, and generates draft reports ready for human review. Or a trading operations agent that adjusts portfolio models based on real-time market signals, without needing constant human input. These agents aren't sitting in isolation. They're embedded into workflows, triggering cross-functional processes and reducing the friction that typically builds up between tools, teams, and data. And because they can run 24/7 without fatigue or context switching, they give small teams the operational capacity of much larger ones - without the organisational drag. Asymmetrical Leverage in Action The real unlock here is asymmetry. Traditional scaling is linear: more people, more output. Agent-first scaling is exponential: more intelligence per task, more value per person. For founders and operators, this is a fundamental shift in how work gets done. Take a UK-based neobank that recently rolled out an internal agent stack to manage financial operations. Instead of adding headcount to reconcile transactions, generate audit trails, and update internal dashboards, they deployed agents to handle these tasks end-to-end. As a result, a finance team of three now operates like a team of ten - not because they're working longer hours, but because the agents are doing the coordination, tracking, and formatting in the background. Or consider a US-based lending platform where customer service agents used to toggle between five tools to resolve one query. Now, an agent sits between those tools, compiles a customer's profile in seconds, drafts the reply, and even pre-fills CRM updates. One team member can now do what previously took three - and they can focus on building relationships, not piecing together data. This isn't just about cutting costs or doing more with less. It's about restoring human attention to where it matters most: judgment, creativity, strategic insight. By eliminating the constant cognitive drain of fragmented systems and shallow coordination work, agent-based infrastructure gives teams space to think, explore, and act with clarity. Why Now? The Tech Has Caught Up If this sounds too good to be true, it would've been - even 18 months ago. But recent advances in large language models, retrieval-augmented generation (RAG), and agent frameworks have changed the game. It's now possible to build AI agents that navigate APIs, evolve through feedback, and reason across a live context map - not as brittle automations, but as strategic actors. Crucially, these aren't brittle rule-based bots that break when the environment changes. The new wave of agents are adaptable. They don't just follow instructions - they understand objectives. That makes them suitable for high-change, high-ambiguity environments like fintech, where requirements shift, tools evolve, and edge cases are the norm. And because many startups are already operating in cloud-native environments with modern APIs and loosely coupled services, they're perfectly positioned to adopt agent-based infrastructure. In fact, it's often easier for an early-stage fintech to build an agent-powered back office than it is for a traditional player to untangle their legacy systems. Rethinking Operational Architecture For founders, COOs, and Chiefs of Staff, the implication is clear: if you're still building operational capacity by adding headcount, you're likely leaving leverage on the table. The question is no longer how many people do we need? - it's what do we want to automate, augment, or offload entirely? That starts with a mindset shift. Designing operations around agents means rethinking your company as an AI-native system. That means codifying your data into structured semantic graphs, enabling cross-agent collaboration, and building feedback loops where agents not only automate but adapt, reflect, and grow - just like a human team would, but faster. It also means building in feedback loops. The best agent-first teams treat their AI systems like new hires: onboard them, train them, review their output, and let them improve over time. This isn't 'set and forget' automation. It's collaborative infrastructure that evolves alongside the business. The reward? An operational stack that scales without ballooning costs or headcount. A company that can punch above its weight in terms of execution. And a team that spends more time solving problems and less time chasing updates or managing handoffs. The Next Fintech Success Stories We're already seeing the early signs of this shift. The most operationally intelligent fintechs - often the ones that look surprisingly lean from the outside - are quietly using agents to do the work of entire departments. They don't brag about it in pitch decks. They don't need to. Their advantage shows up in faster execution, cleaner operations, and happier teams. This doesn't mean people are obsolete. Far from it. But the role of humans in fintech is changing. It's no longer about scaling output through hiring. It's about designing systems that multiply the value of every person you do hire. That's the essence of leverage. And in a sector where margins are tight, competition is fierce, and compliance is non-negotiable, it could be the difference between treading water and building a category-defining business. Conclusion: Build the System, Not Just the Team In fintech, growth has historically been a headcount game. But that era is ending. The companies that succeed over the next five years won't be the ones with the biggest teams - they'll be the ones with the smartest infrastructure. Autonomous agents offer a new path: one where adaptability scales faster than bureaucracy, and intelligence compounds faster than payroll. So if you're building a fintech startup in 2025, ask yourself: are you hiring for leverage - or designing for it? Because the smartest teams aren't growing by the dozen. They're growing by the agent.
Reuters
5 hours ago
- Reuters
Amazon to invest $13 billion in Australia's data center infrastructure
June 14 (Reuters) - Amazon will invest 20 billion Australian dollars ($12.97 billion) from 2025 to 2029 to expand, operate and maintain its data center infrastructure in Australia, aiming to bolster the nation's artificial intelligence capabilities, the company said in a blog post on Saturday. ($1 = 1.5418 Australian dollars)
Scottish Sun
5 hours ago
- Scottish Sun
Gary O'Neil lined up for football return as Championship side prepare to sack boss after disappointing end to season
The Blades are also considering a new director of football WOLF BACK Gary O'Neil lined up for football return as Championship side prepare to sack boss after disappointing end to season Click to share on X/Twitter (Opens in new window) Click to share on Facebook (Opens in new window) GARY O'NEIL tops Sheffield United's wanted list as they look into a change of boss. O'Neil has not worked since leaving Wolves in December but has big admirers at Bramall Lane where Chris Wilder's position is being put under threat. Sign up for Scottish Sun newsletter Sign up 3 Gary O'Neil is being targeted by Sheffield United Credit: Getty 3 Chris Wilder's job is under serious threat Credit: Shutterstock Editorial Wilder, 57, is going to have to handle the new American owners giving him players based on AI data — as we revealed. The experienced gaffer wants a handful of his own targets — but if they do not happen, there will be trouble ahead. Wilder wants experienced players like free agent Jimmy Dunne and old boy Oli McBurnie, while Chris Mepham is another possible arrival. But the new advisers are working on a deal for a mystery Bulgarian right-back, which could spark a major row. READ MORE IN FOOTBALL 'harsh and unfair' Club kicked out of Europe by Uefa in Crystal Palace-style case Wilder signed a new deal in January and the compensation would be huge, a problem that may stretch the new owners' finances. But O'Neil is in line to get the SOS call, although he is also a potential target for other jobs. O'Neil, 42, was not in the market for work after his Wolves exit but is now prepared to take the right job. Former Hull boss Ruben Selles is another live candidate. BEST ONLINE CASINOS - TOP SITES IN THE UK Blades may also name a new director of football, with Bristol City's Brian Tinnion tipped for the post. Sheffield United finished third in the Championship last term. EFL club release 'gorgeous work of art' kit and even rival fans want to buy it Wilder's men led the table at the start of April, but a poor end to the season saw them slip into the play-offs. The Blades comfortably saw off Bristol City in the semi-finals, only to be beaten by Sunderland at Wembley. 3 TRANSFER NEWS LIVE - KEEP UP WITH ALL THE LATEST FROM A BUSY SUMMER WINDOW
