At its simplest, data sovereignty refers to the idea that data is governed by the laws of the country where it is collected, stored or processed. Basically, geography still very much matters, even in a digital world. Data doesn’t just exist in a vacuum – it sits within legal borders, and those borders determine who can access it, regulate it and control it. It’s not exactly straightforward.

Suddenly, Data Sovereignty Matters More Than Ever

For a long time, the internet operated as if data was borderless. Cloud computing only reinforced that idea, allowing companies to store and process information wherever it made the most sense from a cost or performance perspective, but that model is starting to shift.

Governments are becoming far more protective over data, particularly when it comes to national security, citizen privacy and economic value, and for good reason, given the current geopolitical situation. But, at the same time, the rise of AI has transformed data into something far more strategic. Now, it’s just a thing that companies collect kind of on the side – it’s something they compete over and it’s considered a valuable asset. As a result, data sovereignty has moved from a niche legal concept to one of the most central issues in global tech.

Data Sovereignty Isn’t the Same As Data Localisation

It’s easy to confuse data sovereignty with data localisation, but they’re not quite the same thing. Data sovereignty is about legal control and which country’s laws apply to a dataset. And data localisation, on the other hand, is about physical location and where that data is actually stored.

The two often overlap, but not always. For instance, a company could store data in one country while still being subject to the laws of another, depending on ownership or infrastructure. That grey area is where much of today’s tension sits.

The Cloud Adds An Extra Complication

Now, cloud computing has made data sovereignty significantly harder to navigate. Platforms like Amazon Web Services, Microsoft Azure and Google Cloud allow businesses to operate globally from day one. But in doing so, they also blur the lines around where data actually resides and which laws apply.

Data is often distributed across multiple regions, sometimes without companies fully realising it. That raises difficult questions around jurisdiction, access, and compliance – especially when different countries have competing legal claims over the same information.

For lots of businesses, particularly startups scaling internationally, this adds a layer of complexity that didn’t exist a decade ago.

Data Sovereignty Is Becoming a Growing Priority For Startups

What used to be a concern mainly for governments and large enterprises is now firmly on the radar for startups. As more companies build around data (through AI models, analytics platforms or digital services), they’re increasingly being asked to demonstrate where their data lives and how it’s governed. It’s not only a compliance issue. Now, it can influence customer trust, investor confidence and even market access – it’s pretty far reaching.

In some cases, strong data sovereignty practices are becoming a selling point, particularly in regions with strict regulatory environments.

Data Sovereignty On a Broader Level

Now, data sovereignty reflects quite a big shift in how data itself is viewed. It’s no longer just a byproduct of digital activity; it’s an asset, a strategic resource and in many cases, a matter of national importance.

For businesses, that means the rules have changed. Data can no longer just be stored wherever it’s cheapest or most convenient. Now, it actually needs to be handled with an understanding of legal boundaries, political realities and long-term risk.

Because in today’s landscape, where your data sits and who has authority over it is becoming just as important as what you do with it.

The post What Is Data Sovereignty? appeared first on TechRound.

You know that something isn’t quite right. You just can’t quite put your finger on it.

Well, OpenAI has now officially put its finger on it. In mid-March 2026, the company released an update to GPT-5.3 Uniquely designed to reduce what it called “teaser-style phrasing” – the open loops, cliffhanger hooks and high-strung build-ups that had seeped into ChatGPT’s responses. Cases cited in the release notes included phrases like “If you want…”, “You’ll never believe…” and “I can tell you these three things that…”. If those sound familiar, that’s because you’ve definitely been on the receiving end of these ChatGPT responses recently.

Sam Altman, to his credit, was surprisingly candid about it. When a user noted that ChatGPT’s “most distinguishing characteristic is its humanity”, Altman agreed and framed it as a problem, not a compliment. The implication being that an AI known primarily for its warmth and personality rather than its accuracy and directness might be heading off course.

What Changed With ChatGPT – And Why?

Somewhere in the training process, recent versions of ChatGPT were optimised in ways that rewarded engagement over honesty. Praise the user, build anticipation, mirror their enthusiasm back at them. It works brilliantly if your goal is for people to keep chatting, less so if your goal is a straight answer.

This isn’t a conspiracy. It’s a known risk in how large language models get tuned. When human feedback shapes model behaviour, which it does, models learn that flattery and dramatic framing tend to land well so they do more of it. The result, over time, is an AI that has quietly shifted from “useful tool” to “very agreeable pal.”

The broader conversation about whether AI tools are actually serving users or just keeping them engaged has been building for a while. This update is OpenAI acknowledging, in release-note language, that the criticism had merit, indeed.

Why This Actually Matters For Your Business

Here’s where it stops being funny and starts being worth taking seriously.

If you’re a founder or a business owner using ChatGPT for anything that involves judgement – strategy, copywriting, product decisions, customer messaging – you’ve been getting responses from a model that was, at least partially, optimised to make you feel good about your ideas rather than truly pressure-test them.

You ask ChatGPT whether your new product positioning works. It tells you it’s compelling, insightful and really speaks to your target audience. Is that because it’s true? Or because the model has learned that enthusiastic validation gets a better response than “actually, your third paragraph is confusing and your call to action is buried”?

Long-time users have noted that recent versions spend more time praising their questions than challenging them. That’s the kind of thing that feels pleasant in the moment and quietly reinforces confirmation bias over time.

For startups making fast, high-stakes decisions on limited information, confirmation bias is a genuine risk.

The “Quit-GPT” Crowd Had A Point

It would be easy to dismiss the vocal backlash on social media – the “Quit-GPT” communities, the threads about ChatGPT becoming “too human” – as the usual internet noise. But the fact that OpenAI responded with an actual product update suggests it wasn’t just noise.

There’s a reason AI slop has become a real concern across industries that rely on content quality. When AI tools optimise for engagement rather than accuracy, the output gets softer and more eager to please. That’s fine for generating a birthday message, but not so much for a competitive analysis or a risk assessment.

The irony, of course, is that other AI writing tools have faced their own credibility questions in recent years. The pattern is consistent: when the model’s job is to keep you happy, the model keeps you happy. Whether that’s actually useful is a separate debate.

What To Do About It

The bottom line is fairly simple, even if it requires a slight shift in how you use these tools.

Stop asking ChatGPT whether your idea is good, ask it to find the holes in your idea. Prompt it to steelman the opposition, identify weaknesses, play devil’s advocate. A model that’s been tuned toward agreeableness will still push back if you explicitly ask it to – it just won’t volunteer the criticism if you don’ t prompt it to.

And be conscious of the flattery. When ChatGPT opens with “what a great question” or tells you your draft is “really compelling”, that’s not editorial feedback. That’s tone-setting. The actual content that follows is usually more useful than the preamble suggests, but only if you’re reading past its opening remarks.

OpenAI has made a start. The responsible development of AI tools depends on exactly this kind of honest self-correction. But the update won’t solve everything overnight, and the underlying incentives that created the problem haven’t disappeared.

The most useful thing an AI tool can do for your business is tell you something you didn’t already want to hear. Worth keeping that in mind next time it tells you your strategy is absolutely brilliant.

The post ChatGPT Has Been Flattering You And It Could Be Costing Your Business appeared first on TechRound.

To recap, in the proposal, a company can register within 48 hours, fully online, at a maximum cost of €100 and with no minimum share capital. The Commission says founders will submit their information once through an EU level interface connecting national registers. A central EU register will follow in a second phase.

President Ursula von der Leyen said: “Europe has the talent, the ideas and the ambition to become the best place for innovators. Yet today, European entrepreneurs who want to scale up face 27 legal systems and more than 60 national company forms. With EU Inc., we are making it drastically easier to start and grow a business all across Europe. Any entrepreneur will be able to create a company within 48 hours, from anywhere in the European Union, and fully online. This crucial step is just the beginning. Our goal is clear: one Europe – one market – by 2028.”

For digital first finance companies, time and cost are decisive. A payments startup in one member state could register once and operate in multiple countries under one corporate structure, instead of recreating itself country by country.

How Could It Changr Access To Capital And Markets?

 
Fintech depends on venture capital and public listings to grow. The Commission says EU Inc. will make the transfer of shares much simpler, and remove mandatory intermediaries in share transfers. It will also allow member states to grant EU Inc. companies access to stock exchanges.

The factsheet describes EU Inc. as “Attractive for investors” through simple transfer of shares and stock exchange access. For finance companies that grow quickly and require repeat funding rounds, fewer formalities can lower transaction costs and shorten fundraising timelines.

The Communication published with the proposal outlines changes connected to the Savings and Investment Union. These cover a review of the European Venture Capital Funds Regulation and a possible revision of pension fund investment rules. Pension capital represents long term funding that fintech companies often court.

On tax, the proposed Head Office Tax system would let small and medium sized enterprises apply the tax rules of their home country. The Business in Europe Framework for Income Taxation initiative would create a single legislative framework for corporate taxation. For cross border finance companies, tax predictability can influence where they base operations.

And What About Rules Around Insolvency And Talent?

 
Finance and fintech compete hard for engineers, compliance staff and product teams. EU Inc. companies will be able to set up EU wide employee stock option plans. Stock options will be taxed when income is generated and sold, which can make equity packages more appealing to staff.

The Commission also plans to explore 100% cross border telework for startups and scaleups under the forthcoming Fair Labour Mobility Package. That could help fintech hire in different member states without relocating entire teams.

On risk, the factsheet describes EU Inc. as “Low risk” through simplified and digital insolvency for startups. Founders will have access to fully digital liquidation procedures and simplified insolvency rules. That gives entrepreneurs a more predictable restart option if a product fails.

National employment and social laws will apply in full. Member states are encouraged to set up specialised courts to handle EU Inc. company law cases. For the finance sector, legal certainty may prove just as important as the 48 hour registration promise.
 

How Do Experts Think The EU Inc Will Impact Fintech?

Our Experts:

• Jeremy Brown, Investment Principal, Anthemis
• Laurent Descout, CEO and Founder, Neo
• Jana Zdravecka, CFO, INFINOX
• Arthur Azizov, Founder and Investor, B2 Ventures
• Alpesh Patel, Strategic Partnership Director, Cartex
• Rob Rooney, CEO, Hyperlayer
• Pedro Rychter, Vice President, AVP
• Fabio Coco, Partner, ADVANT Nctm
• Becky Simms, CEO, Reflect Digital

Jeremy Brown, Investment Principal, Anthemis

 
“For fintechs, EU Inc is directionally right, but it only solves part of the problem. While it could reduce legal and administrative friction when expanding across borders, more work will need to be done to address fragmented licensing regimes, uneven capital markets and different go-to-market conditions, all which continue to be barriers to scale quickly.”

Laurent Descout, CEO and Founder, Neo

 
“As the founder of a Barcelona-based fintech, I see firsthand just how nightmarish scaling a fintech in Europe can be. You choose to incorporate in one country, and then face 27 different approaches to company laws, taxes and administrative processes if you choose to scale across the EU. By directly tackling these pain points and simplifying the cross-border scaling process, EU Inc will allow startups to tap into new pools of capital, extend their investor base and better compete with wealthier global giants who can easily absorb the cost of setting up many times over.

“However, for all the benefits it promises in theory, success will be determined on how it works in practice. Centralisation on paper is not enough. Instead, supervision, licensing, and enforcement needs to be genuinely applied in a uniform manner across all member states.”

Jana Zdravecka, CFO, INFINOX

“The EU Inc proposal is a meaningful step towards reducing the administrative complexity of building a business across Europe. However, simplifying company formation addresses only one part of the challenge.

In practice, the greater barrier to scaling businesses across the region has been how capital moves, and how easily investors can access opportunities across different markets. Despite the concept of a single market, fragmentation in investor access, onboarding requirements and regulatory interpretation continues to create friction in cross-border participation.

While INFINOX does not operate within the EU, we work with a global client base and see first-hand how differences in market structure and accessibility influence where capital is deployed. The efficiency of European markets, and the ease with which capital can move across borders, plays a significant role in how competitive the region is in attracting and retaining high-growth businesses.

From a fintech perspective, improving the accessibility and connectivity of capital markets is critical. If Europe is to support the next generation of companies, it must ensure that investors, both institutional and retail, can engage with opportunities seamlessly across the region.

Ultimately, the success of this initiative will not be measured by how easily companies can be formed, but by how effectively they can be funded, scaled and accessed within a truly integrated market environment.”

Arthur Azizov, Founder & Investor, B2 Ventures (B2BROKER and B2BINPAY)

“With my experience founding fintech companies with operations in multiple jurisdictions, for me, Europe has always felt harder than it should be. Every step comes with an extra layer of structure, whether new entities, legal setups, or different interpretations of the same rules, slowing down decisions, which in other markets would take days.

“EU Inc will help to start faster, acknowledging a problem that founders, like me, have been dealing with for years. Simplifying the establishment of companies allows a more unified base to operate from. And for fintechs, this makes a huge difference.

“Most infrastructure businesses need to scale across borders from day one. With the current setup, companies have to think locally before regionally. This goes against how modern financial products are designed.

“I do not think founders will overestimate the impact. The real friction in fintech is licensing, compliance and the need to deal with different regulators and approaches.

“In order for Europe to gain more competitiveness, it needs to go further. Until financial regulation becomes as coordinated as company formation, Europe will still feel fragmented when it matters most.”

Alpesh Patel, Strategic Partnership Director, Cartex

“EU Inc is directionally right. Founders across Europe have been asking for this for years. Faster incorporation, less fragmentation, and the ability to scale across borders without rebuilding structure in every country. That problem is real, and this proposal targets it directly.

“If implemented properly, it will remove a layer of friction for early-stage fintech’s. You get faster market entry, cleaner corporate structures, and a more consistent framework for hiring and equity. That matters for founders and investors. It improves how capital flows and how quickly companies can execute.

“But let’s be clear. This does not solve the core regulatory problem. Licensing, acquiring, and local compliance requirements will still sit at the country level. For fintech, that is where most of the complexity and cost sit today.

“So, the upside is clear, but execution is everything. If member states align and adoption is real, this can strengthen Europe’s position and improve investor confidence. If not, it risks becoming another well-intended framework that falls short in practice.”

Rob Rooney, CEO, Hyperlayer

“This is a serious proposal to remove friction at scale. A 48-hour setup and a single rulebook across EU markets goes straight at the fragmentation that’s slowed European startups for years. However, while friction is certainly an issue, it doesn’t address the much bigger challenge facing both UK and EU start-ups: access to capital. The gap in unicorn numbers reflects that. Without deeper pools of funding to support innovative companies as they scale, that’s unlikely to shift significantly.

“For UK fintechs, it makes the comparison clearer. The UK remains a strong base, but if the EU can offer that level of simplicity across multiple markets, more companies will factor it in earlier when deciding how to structure and scale. The UK’s response needs to be practical: improve access to capital, keep systems simple, and continue attracting talent. Founders don’t overthink it: they build where things can scale fastest.”

“Hyperlayer, which is UK based with operations in US and Asia PAC, is an innovation layer for modern banking that enables banks to launch intelligent financial products in weeks without replacing their core systems.”

Pedro Rychter, Vice President, AVP

“As we have seen with our portfolio companies, the fastest growing and highest quality fintechs in Europe are those able to scale simultaneously across multiple countries, tapping into the entire profit pool of Europe instead of focusing on niche local markets.

“However, in practice this is much less seamless than scaling in the US because of the patchwork of local specificities – for instance, we’ve seen that one big barrier to recruit and retain talent in other countries was the lack of a unified ESOP framework. A German employee exercising French stock options (but not selling them for cash) would still be required to pay taxes, which makes hiring top talent in Germany naturally harder for a French company.

“By partially removing those intra-EU roadblocks, the EU Inc is a major win for the fintech ecosystem and a big step in the direction of unlocking growth and innovation and Europe, a continent still punching below its weight then it comes to technology.”

Fabio Coco, Partner, ADVANT Nctm

“The Commission unveiled the EU Inc. proposal on 18 March: it falls short of what the financial sector urgently needs to accelerate innovation and it does not offer any form of ‘fintech passport’ or a genuine ‘single license’ for financial services.

“At this stage, the ‘EU Inc.’ proposal, seems to be a missed opportunity for innovation in the financial sector: the absence of an EU-wide regulatory sandbox or ‘safe harbor’ provision is a notable omission. Such an environment is vital for testing innovative financial services under a reduced regulatory burden without compromising market integrity. By deferring to existing national frameworks, the proposal misses a critical opportunity to lower the high barriers to entry in the banking, insurance, and financial sectors. This lack of centralised ambition risks curbing the competition and innovation the initiative was intended to foster.

“In contrast, the progress made by the Italian Supervisory Authorities serves as a robust benchmark. Facilitated by a forward-thinking legislative approach, Italy successfully implemented a national regulatory sandbox. This controlled environment allows supervised entities and Fintech operators to test technologically innovative products and services within the banking, financial and insurance sectors for a defined period – a model that provides the clarity and flexibility currently missing from the broader EU proposal.”

Becky Simms, CEO, Reflect Digital

“One of the biggest challenges in the global race for AI talent is actually defining what ‘AI talent’ means. For some organisations it refers to highly technical specialists who can build models and integrate AI into existing systems. For others, it simply means hiring people who are AI-native in how they work and use AI tools in their day-to-day roles. We’re still very early in this shift, and that definition isn’t always clear.

“That ambiguity is where the hiring risks begin. Many organisations are recruiting for AI roles for the first time without having the internal expertise needed to properly benchmark candidates’ skills. At the same time, boards are pushing companies to move quickly so they don’t fall behind.

“The challenge is that AI is evolving at such speed that what looks like the right skill set today could look very different in three months’ time. If businesses rush into hiring without clear frameworks or technical oversight, they risk bringing in people who don’t have the depth of expertise required, or steering their AI strategy in the wrong direction.

“In many cases, the smartest move is to bring in a trusted advisor or technical specialist to support the recruitment process. That ensures organisations aren’t just reacting to the hype cycle, but building AI capability in a way that’s genuinely strategic and sustainable.”

The post What Will The EU Inc. Proposal Mean For The Fintech Industry? Experts Discuss appeared first on TechRound.

Research from Studio Graphene, based on a survey of 500 UK managers, directors and C suite executives carried out by Censuswide, found that 78% of businesses use AI in some form. In mid sized firms with 100 to 249 staff, that number comes up to 85%.

Usage is high. Financial return is low.

Why Are Most Firms Not Seeing A Return On AI?

 
Only 31% of businesses using AI said they have seen a positive return on investment from their spending. That means nearly seven in ten have not made clear financial gains so far.

The same survey found that 18% said their AI projects did not deliver the benefits they expected. Another 16% said it was too early to judge the outcome. When you add those groups together, more than a third of AI users either feel disappointed or cannot tell if the technology is paying off.

The issue may start with the fact that there’s no direction. Just 41% of AI users said they have defined what “success” looks like when they bring in AI tools. In mid sized companies, which lead usage at 85%, only 46% said they can define success. If leaders cannot describe what winning looks like, measuring ROI becomes difficult.

Boards that approved spending now face pressure. Money has gone out, but proof of value has not always come back in.

Are Businesses Spending Before Setting Targets?

 
Ritam Gandhi, director and founder of Studio Graphene, believes many companies rushed in without setting firm criteria for returns.

He said: “Many organisations are at a critical point in their AI journey. Adoption has skyrocketed in the past year, particularly among mid-sized businesses, but our research clearly shows just how much progress is required for AI projects to be successful.

“There has been a rush to adopt AI amidst huge hype and a proliferation of new tools – this is certainly true of private equity-backed mid-sized companies looking to AI for automation, scalability and competitive edge.

“The problem, however, comes when AI is deployed without first defining where it sits within workflow, the decisions it’ll inform, the processes it’ll support, and the criteria for measuring success – often teams haven’t agreed whether AI is meant to save time, improve decision quality, reduce risk, support growth or all of the above.

“It’s a really important issue that threatens progress. Without defining these things, building a long-term business case for AI and realising its value will be difficult. At board level, frustration will grow without a clear picture of how and why AI is being used, and to what effect. It underlines the need for rigorous planning for any AI transformation project, not just in selecting the right tools, but in defining the broader strategy, implementation and criteria for success.”

His comments go straight to ROI. If a company does not decide in advance how AI should save money or increase revenue, it cannot properly judge performance later.

What Does This Mean For Business Budgets?

 
The survey of 500 UK businesses shows a divide between enthusiasm and earnings. 78% use AI. Only 31% report positive ROI. That difference will influence how boards view future spending.

Private equity backed mid sized firms appear keen to bring in AI for automation and scalability, according to Gandhi. Those investments often come with expectations of fast financial returns. When those returns do not appear, patience can become thin.

AI is now common in UK offices. Profit from it is far less common. The next round of spending may depend less on hype and more on proving that AI can bring pounds back into the business, instead of simply adding new software into the building.

The post New Data Shows That Only 31% Of UK Firms Report Positive ROI From AI appeared first on TechRound.

The current tension in what we’ll call “The Days of our AI Lives” is between Microsoft, OpenAI and Amazon, and word on the street is that Microsoft isn’t particularly sweet on OpenAI anymore.

I can’t help but read the situation like the tale of a sordid love affair – one that started with great excitement and possibility but seems to be ending in distrust and betrayal. And to lean further into the metaphor, a potentially messy divorce, too.

Indeed, the dispute is centred on something far less dramatic on the surface, but arguably far more important that is, who gets to control where OpenAI’s AI actually runs?

The Original Relationship Was Built On Exclusivity

At the beginning of this partnership, Microsoft and OpenAI were very much aligned.

Microsoft poured billions into OpenAI in its early stages, and, in return, became its exclusive cloud provider. That meant OpenAI’s models, including GPT, would run on Microsoft Azure, giving Microsoft a powerful edge in the cloud and AI race.

It was a truly mutually beneficial setup – OpenAI got the compute it desperately needed to scale, and Microsoft got front-row access to some of the most advanced AI models in the world.

Think of it less as a casual partnership and more as a fully committed alliance – Jedi-level trust, if you will.

Cracks Begin To Show

But as with most alliances in tech, things didn’t stay simple.

As demand for AI exploded, so did the need for more compute, more flexibility and fewer constraints. Relying on a single cloud provider, even one as dominant as Microsoft, started to look less like a strength and more like a limitation.

Over time, the agreement between the two reportedly evolved. While Microsoft still retained key rights (particularly around access to OpenAI’s models via APIs), there was growing room for interpretation when it came to new types of AI systems. And interpret, OpenAI did.

Unsurprisingly, that’s where things started to get a little complicated.

Amazon Enters the Chat

Over time, the terms of the initial deal have shifted somewhat, with lines becoming blurry. It seems like Microsoft has had eyes on the situation and has (mostly) been part of these conversations, allowing these things to happen. But now, it seems like OpenAI and Altman have pushed the boundaries a little too far.

OpenAI has reportedly struck a major deal with Amazon Web Services, using its infrastructure to support more advanced AI systems – particularly those that go beyond simple, stateless interactions.

These are the next generation of AI tools – systems with memory, autonomy and the ability to act more like agents than chatbots.

From OpenAI’s perspective, this is a logical move. More partners mean more compute, less dependency and greater room to grow. But, from Microsoft’s perspective, it’s a loss of control and a potential breach of trust.

Here’s Why Microsoft Is Unhappy and Why They May Have A Legal Leg To Stand On

At the heart of the dispute is a simple but critical question – what exactly was Microsoft paying for?

If the original deal guaranteed Azure exclusivity, then moving significant workloads to AWS could be seen as undermining that agreement. In my opinion, quite clearly so.

But if that exclusivity only applies to certain use cases like API access, then OpenAI may be well within its rights to expand elsewhere. Perhaps not one made in good faith, but not necessarily a legal issue.

That ambiguity is what’s fuelling the tension.

And this is where the “Star Wars” analogy starts to feel less like a stretch and more like a blueprint. What began as a powerful alliance is now edging toward something far more adversarial – with control, power and territory all up for grabs.

OpenAI has already found itself at the centre of controversy (once again) with the whole Anthropic, defence tech debacle with the Pentagon and the US government. A move that many would argue resulted in a significant loss of public trust, with some even moving to boycott ChatGPT and more entirely.

If we’re sticking to the “Star Wars” universe, this would be the whole “River of Lava” lightsaber duel between Anakin (ie. OpenAI) and Obi Wan (Anthropic) in “Revenge of the Sith” – the one that ultimately ends with Anakin crossing over to the Dark Side and becoming Darth Vader.

And this whole OpenAI-Microsoft conflict?  That’s got to be the iconic “No, I am your father” scene in “The Empire Strikes Back”. A betrayal for Luke (Microsoft), but it’s also the beginning of a new, pretty dire time for the Rebel Alliance – a loss of power and control.

So, will the AI conflict follow the same trajectory of the “Star Wars” narrative? Is Microsoft going to take a big hit here and fight from the back foot, while OpenAI becomes increasingly powerful and seizes even more control? Or, will Microsoft flip the script, win the case and keep OpenAI (also read, “the Dark Side”) at bay?

The AI race may not be quite as dramatic as the “Star Wars” universe, but the conflict, betrayals and constant fight for control is pretty reminiscent of what we’re seeing at the moment.

Because in today’s AI race, it’s not just about building the best models.

It’s about controlling the galaxy they run in.

The post In A Galaxy Not So Far Away, The “Empire” Strikes Back – Microsoft And OpenAI Heat Up The Cloud AI Wars appeared first on TechRound.

The acquisition makes strategic sense once you understand what Mastercard is actually buying. BVNK isn’t just a crypto company – it’s infrastructure. Founded 2021, it operates across 130 countries, holds EU CASP licensing and SEPA access, and has processed around $30 billion annually, according to Mastercard. It enables seamless conversion between fiat currencies and stablecoins at scale. For a global payments network looking to move beyond traditional rails, that’s a compelling proposition.

The timing plays a role as well. This deal follows Stripe’s acquisition of Bridge in 2025, and sits within a broader race among payment giants to own crypto infrastructure before it becomes table stakes. Stablecoin transaction volume hit $350 billion in 2025, according to industry data – the evidence suggests that these are no longer niche instruments. As Mastercard’s chief product officer Jorn Lambert put it, the deal is about bringing “tokenized money’s benefits to real-world use.” The UK’s own investments in next-generation financial infrastructure suggest policymakers are watching the same horizon.

The question this deal forces is a bigger one: are stablecoins crossing over from a crypto conversation into a mainstream finance one? The biggest fintech funding rounds of 2025 suggested the market was already moving in this direction. The BVNK deal may be the moment it becomes undeniable. To weigh in on this, we asked five industry experts who work at the sharp end of this shift.

Our Experts:

• Melvis Langyintuo: Executive Director, Canton Foundation
• Charles Buckworth: Partner at RPC
• Simon Jones: Co-Founder and CEO at Sumvin, Inc.
• Shantnoo Saxsena: CEO and Founder, Encryptus
• Bjarni Thor Sigurdsson: Chief Commercial Officer, PAYSTRAX
• Sid Powell: CEO and Co-Founder, Maple

For any questions, comments or features, please contact us directly.

Melvis Langyintuo, Executive Director, Canton Foundation

“This deal exposes a deep structural tension. Stablecoins are scaling fastest in payments, but payments are the least complex part of financial markets. The harder problem is what happens next when that same digital money needs to interact with collateral, securities and financing workflows. If those layers are not interoperable, liquidity doesn’t compound and instead fragments across parallel systems.

“The industry still treats interoperability as a downstream integration challenge. In practice, it is upstream market design. You are aligning how different assets move and behave, how transactions are sequenced and, critically, who has visibility into each step. That cannot be solved bilaterally or retrofitted through APIs. It requires shared rules from the outset.

“This is where vendor-neutral governance becomes critical. Without it, every network optimises for its own perimeter – standards, data models and liquidity pools. The result is not competition on efficiency but competition on fragmentation.

“In our view, how stablecoins will plug into a synchronised financial system, or proliferate as disconnected balance sheet extensions, is critical. The difference determines whether we get incremental gains in payments or a step change in how global markets function.”

Charles Buckworth, Partner, RPC

“This acquisition is another sign that we are moving past a battle of ‘fiat against crypto’ to a payments ecosystem of co-existence or perhaps even symbiosis.

“Innovation in payments only gains traction if those innovations sit within a trusted framework which reduces or removes friction in the user experience. The adoption of stablecoin has been constrained to date by the absence of the trust, governance frameworks and incentives needed for widespread adoption, together with an often-poor user experience and journey.

“The market is now responding to these constraints, with increasing focus and investment from major payment and fintech providers. The Mastercard/BVNK deal allows Mastercard to connect and integrate stablecoin into a trusted network with global reach, wrap-around governance frameworks and longstanding regulatory relationships. This transaction signals increased institutional and consumer adoption in crypto and stablecoins and perhaps also that future innovations in fintech, including crypto and stablecoins, are more likely to be integrated into existing payment networks, rather than replacing them entirely.”

Simon Jones, Co-Founder and CEO, Sumvin, Inc

“Mastercard’s reported $1.8billion acquisition of BVNK is a defining moment – and a clear signal that stablecoin infrastructure has moved firmly into the mainstream financial conversation. When a network of Mastercard’s scale makes this kind of commitment, it doesn’t just validate the technology; it accelerates the entire ecosystem.

“What this deal really tells me is that the era of institutions sitting on the sidelines is over. Stablecoins offer programmable, near-instant, low-cost settlement that legacy systems have struggled to match – and the smartest players in tradfi are now moving to get ahead of that curve.

“For all those building on blockchain infrastructure, this is meaningful validation – but complacency would be a mistake. Institutional capital moving in at this scale compresses the window to carve out a defensible position. The race now isn’t about whether stablecoin-backed infrastructure gets built, it’s about who controls the execution layer on top of it.”

Shantnoo Saxsena, CEO and Founder, Encryptus

“This may well be the most significant payments deal of the decade, and the clearest signal yet that stablecoin infrastructure has become critical plumbing for global finance.

“What makes this significant is the scale of conviction behind it. Mastercard could have partnered, invested or acquired a smaller player. Instead, they paid a premium for a company that has built genuine enterprise-grade stablecoin rails. That tells you everything about where they see payments heading.

“For fintechs and crypto-native businesses operating in regulated markets, this validates the trajectory we have been building towards. The companies that will define the next phase of fintech are those already operating at the intersection of compliant infrastructure and digital assets. The line between traditional finance and crypto infrastructure is disappearing – and deals like this are accelerating that convergence.

“With Stripe and now Mastercard moving decisively on stablecoin infrastructure, the question is no longer whether traditional finance will adopt blockchain settlement – but how quickly regulated players can scale it before unregulated alternatives fill the gap.”

Bjarni Thor Sigurdsson, Chief Commercial Officer, PAYSTRAX

“Mastercard’s investment in BVNK signals that stablecoins are moving from the fringes to the financial mainstream. And the regulatory groundwork is already being laid – the UK is already building toward a formal stablecoin regime. The FCA has prioritised stablecoin payments for 2026 and the Bank of England is consulting on systemic stablecoins.

“The opportunity here doesn’t lie in replacing cards or bank accounts. It’s in the settlement and money-movement layer underneath – including cross-border flows, merchant settlement and treasury operations. Stablecoin rails can reduce delays, extend settlement beyond banking hours and bring much-needed transparency across the payment lifecycle.

“The broader shift we’re seeing is that payments are becoming programmable, meaning faster disbursements, more automated reconciliation, and cleaner connections between fiat, cards and digital assets. As regulation becomes clearer, we’re likely to see stablecoins increasingly win where speed, cost and global reach matter most.”

Sid Powell, CEO and Co-Founder, Maple

“This deal isn’t just a bet on stablecoins – it is actual confirmation. Mastercard’s move into BVNK shows that stablecoins have moved past the line from being a crypto-native tool to becoming a core part of global financial infrastructure. What this deal really signals is that the market has moved past theory and that the rails matter.

“Large payment players aren’t adopting stablecoins because they are trendy. They are a faster, more reliable way to move money globally, with settlement certainty that legacy systems simply can’t match.

“At Maple, we see this demand every day from our institutional clients who require speed and certainty minus the friction of traditional finance. The race now will be centred on who builds the best products – and that’s where the real competition is just getting started. For incumbents, that means taking this shift seriously. For startups, it creates real room to win by building faster and with fewer constraints than legacy systems allow.”

What Comes Next?

What’s striking about these expert views is the consensus on direction. The debate isn’t whether stablecoins become part of mainstream financial infrastructure – but whether their rules and safeguards can keep pace with the speed of adoption. As fintech hiring and investment continues to grow in the UK, the talent and capital are clearly betting on the same outcome.

For London’s startup ecosystem, the BVNK deal is also a significant moment in its own right. A four-year-old company built here just became one of the most valuable fintech acquisitions of the year.

That’s not a footnote – it’s a signal about where the world’s most ambitious fintech builders are still choosing to work.

For any questions, comments or features, please contact us directly.

The post Experts Comment: Is Mastercard’s £1.4B Investment In A London Stablecoin Startup A Sign Of Fintech’s Next Frontier? appeared first on TechRound.

From AI-powered accounting platforms and autonomous hiring agents to gamified learning tools and green energy automation, the startups emerging from Denmark are diverse, ambitious, and increasingly hard to ignore. Here are the top SaaS startups in Denmark you need to watch.

How is SaaS Changing Denmark’s Tech Landscape?

By replacing traditional, locally installed software with cloud-native, subscription-based solutions, SaaS is quickly changing Denmark’s digital scene, and the industry is predicted to expand. Government initiatives, a strong demand for AI-integrated products, and the SME sector’s rapid digitalisation are driving this evolution.

What Challenges Are Facing SaaS Startups in Denmark?

SaaS entrepreneurs in Denmark face a unique mix of obstacles, from fierce talent competition to regulatory barriers, despite operating in a highly digitalised and established sector. High operating costs, stringent compliance standards, a severe lack of IT and AI expertise and scaling hurdles are some of the main obstacles.

The Top SaaS Startups in Denmark

With the expected growth and the growth the SaaS scene has already seen in Denmark, more businesses are opening their doors, specifically SaaS businesses. Here’s a list of the top SaaS startups in Denmark:

1. Light

The top AI accounting platform is called Light. They combine all of your international financial activities onto a single automated platform. Automate bookkeeping, financial reporting, AR, and AP across all of your organisations to avoid the confusion of several systems.

2. Kiku ApS

Kiku is an AI firm based in Copenhagen that offers “Sara,” an autonomous AI-powered agent, to transform high-volume hiring for frontline employees. It reduces hiring time and expenses for sectors including retail, logistics, and healthcare by automating the process of identifying, screening, interviewing, and shortlisting applicants in more than 100 languages through chat, audio, or video around-the-clock. 

3. Contribe

Online shops can include charitable donations straight into the checkout process with the aid of Contribe, an impact technology SaaS provider. The platform increases customer loyalty and conversion for merchants by allowing customers to give a portion of their purchases to charitable causes at no additional cost. 

4. AI-Energy

The future of energy systems operation and design automation is provided by AI energy. They create technologies that automate solar (photovoltaic) plant planning, optimise energy market bidding, and run a portfolio of energy and power plants.

5. Planet Peanut

A gamified learning tool called Planet Peanut transforms maths courses into cooperative contests. After logging into the mobile app, students work through curriculum-aligned tasks in a fun world that is populated by their own avatar. Students participate in time-limited competitions as teams, within their class or school, rather than working alone. 

The post Top SaaS Startups in Denmark to Watch appeared first on TechRound.

In simple terms, we help brands turn email into a performance channel for acquiring new customers, not just for retention.

Most brands think of email as a CRM tool, but we flip that on its head. Using our data from 17 million opted-in UK consumers, we make it possible for brands to reach completely new audiences in their inbox and beyond.

We do this by understanding your ideal customer through Inbox Intelligence – insight into real consumer behaviour based on what people are engaging with in their inbox. From there, we build high-performing lookalikes and make those audiences targetable across email, programmatic, social, and even direct mail.

Our superpower is scale and intelligence – we can reach one-in-three UK consumers directly in their inbox, while bringing a completely new, privacy-safe channel into the acquisition mix.

We’ve built four products around this:

MirrorMail – turns your best-performing CRM emails into acquisition campaigns that reach 17 million new consumers

Inbox Extend – connects your email audiences with social, programmatic and direct mail for true omnichannel targeting

Optivo – our latest addition to the suite enables brands to retarget anonymous website visitors by email, even before they reach the basket

Customer+ – grows your first-party subscriber base with high-intent, opted-in users ready to convert.

In short, we’re redefining what’s possible with email. We’re helping brands acquire customers more intelligently, using identity rather than cookies.

How did you come up with the idea for the company?

Back in 2014, email was murky. It feels weird to say it, as this doesn’t exist today, but you could buy consumer data on a CD-ROM. In fact, there were two options – buy the data yourself and hope it was okay, or use a provider and blindly pay them, with no transparency on results. Meanwhile, in an adjacent space, programmatic buying was in full swing, which made it easier and more transparent to buy data.

We built esbconnect as we knew email delivered a better return than display, but lacked the transparency and trust. So, we figured we would build a transparent buying platform for email, where people could see the results in real time and select the audiences they wanted to target. This was the Marketplace, and to this day, this is still our most popular product – powering our MirrorMail product.

However behind the scenes, we knew email’s real power was its ability to be an ID, so we have been eating all the cookies in sight for years, quietly waiting for the rest of the industry to catch up and realise it’s a genuine superpower. And now it has, and email fingerprinting now powers many of our scaling products, including Optivo, Ideo and Inbox Extend. All of these make it possible to identify a person online and tie their online identity to an email, enabling omnichannel journeys.

Tell us about your core product or service?

Broadly, it is using email to power acquisition. At our heart, we have an audience of 17m, and that is an audience we know deeply. We know their postal address, whether they have children or pets, and we see what they are doing in their inboxes – who they are clicking on and buying from. We call this data Inbox Intelligence.

Inbox Intelligence powers our ability to offer truly omnichannel journeys. A brand can suppress its current customers, and build a lookalike of its next best customers, targeting them in social, programmatic, postal and in their inbox.

Of our core two products, one is MirrorMail – the ability to use email like Meta to acquire new audiences – select your audience, upload an email creative, and we handle the delivery and optimisation, driving you leads and sales. The other is Optivo – it’s the step before cart abandonment, enabling you to retarget with an email the 90% who never reach the cart and leave your site, encouraging them to subscribe/purchase.

What most excites you about the adtech industry?

It is a cliché but it is AI – but really the ability for websites to become a very personal space for an individual, especially if you think down the line to how the metaverse and VR could play into it. But imagine, you get driven to a website and instead of it being a search box, it’s an avatar of your personal shopper – they remember what you bought before, what your preferences are and they have curated a personalised selection of products that you can virtually try them on and purchase. That excites me.

And what excites me more are the ways brands can move towards this by growing their CRM base. The more you can connect data back to an individual, the better your models and the more personalised the experience. By owning the email address, you can control that conversation and educate them on the value exchange of providing more data. First-party data is king and you should be focused on partners that can help you grow it.

What has been the biggest challenge you’ve had to overcome along the way?

Education – it is easy to spend on Meta, on Google – nobody is going to fire you for doing that. But email isn’t thought of as an acquisition channel, so we continuously have to educate people about its value. Then, even when we do convince people, where do we sit? The CRM team? Or paid media? Or brand? Or performance?

What is your number-one piece of advice for aspiring entrepreneurs?

Well… I’ve got a few. First, plan to fail. Second: your gut knows best in hiring and firing. If you think someone isn’t working out, make that decision quickly. Then, get a good accountant, as cashflow is king. And finally, know your weaknesses and hire for that.

What can we hope to see from esbconnect in the future?

Retargeting is going into its third wave and Optivo from esbconnect will be pivotal for that – when people think of retargeting site abandoners, they will think Optivo.

The post A Chat With Suzanna Chaplin, Founder And CEO, esbconnect, On The Future Of Acquisition And How Email Is The New Identity Layer appeared first on TechRound.

This is more than just improvement; it is a complete reformation of traditional sports stadiums.

What Exactly Is a Smart Stadium?

A smart stadium is a sports venue in which the most advanced technologies are used in every aspect of the venue. They deliver high speed connectivity, AI-powered crowd management, biometric entry, cashless payments and the most advanced technologies to create a seamless and interactive experience to punters.

What Is an Example of a Smart Stadium?

There are strong contenders around the world, such as The Santiago Bernabeu, the Mercedes-Benz Stadium in Atlanta and the Tottenham Hotspur stadium in London. As a smart stadium exemplar, Allegiant Stadium in Las Vegas stands out.

Why is the Allegiant Stadium in Las Vegas Unique?

Inaugurated in 2020, the Las Vegas Raiders NFL team stadium cost US$ 1.9 billion. It was designed and built in three years. It has a 65,000 seat capacity and is considered the most advanced football stadium in America. It has set new standards in football stadium construction and has changed the way spectators experience stadiums.

Features of the Allegiant Stadium

It features over 2,500 Samsung screens, a mobile natural grass field that slides in and out of the venue for football over a concrete floor used for concerts and advanced technology designed for entry and concession access, broadcasting and fan engagement.

VP of Technology for the Raiders Heading to Australia Ahead of 2032 Olympics

Matt Pasco, VP of Technology for the Raiders, has led the design of the Allegiant Stadium and will be in Australia at the end of the month to speak at the inaugural Future of Sport Conference at QUT in Brisbane. This will be the first opportunity for Pasco to speak about the entire process of the stadium and how it is evolving. This topic is especially pertinent to Brisbane as it is the future home of the 2032 Summer Olympic and Paralympic Games and will require advanced stadium and sporting facilities.

10 Amazing Features of a Smart Stadium

From facial recognition for entry, to AI crowd management and personalised mobile apps, smart stadiums have a huge range of interesting smart features to look at:

1. Facial Recognition Entry

No Tickets, No Phone, No Problem. Facial recognition technology scans the fan’s face to identify them in mere seconds. The Cleveland Browns were the first to utilise this technology with Wicket’s Express Access System. It allows entry in two seconds and clears gates 10 minutes faster than the traditional systems which helps clubs save financially.

2. AI-assisted Crowd Management

The halftime toilet queue may be the single biggest unsolved problem in sports. Smart stadiums are finally solving this problem using artificial intelligence and real-time sensor data. They are monitoring crowd density throughout the venue and directing fans away from congested areas. IoT sensors and cameras continuously track crowd behaviour and stream data to systems which can dynamically reroute foot traffic, open new concession stands and notify staff about possible bottleneck situations.

3. Connectivity: 5G and Wi-Fi 6

Stadium Wi-Fi has always been terrible and has been a common joke, but smart stadiums are resolving that issue with the use of private 5G networks and Wi-Fi 6. This combination allows them to accommodate the high demand of internet usage to seamlessly provide stadium-goers with internet access. An example is Wembley Stadium, where Wi-Fi 6 and 5G networks were integrated so that people can use the internet without issues during major events.

4. Augmented Reality Experiences

A genuine disappointment of being in a stadium in comparison to watching at home is the absence of the broadcast layer that has the replays, stats and analysis. With augmented reality, this gap is being closed. Augmented Reality enables a fan to see overlays of real-time data via their smartphones or wearables.

5. Mobile App Technology Solutions

Smart stadium apps are no longer simply digital match day programs; the most advanced apps are becoming all-encompassing fan experiences. Mobile ticketing, seat upgrades and food ordering are all features of these apps. They facilitate the most streamlined and organised user experiences through interactive navigation and real time game updates including stats, team news and strategic loyalty reward programs.

6. AI Referee Technology 

Smart stadium features offer the first generation of AI-assisted officiating technologies that can make decisions in real-time. Data from Premier League shows that referee decision accuracy improved from 82% to 96% after the introduction of VAR; that’s a massive improvement in a sport where one decision can change everything.

During the 2022 World Cup, ball sensors provided real-time data to the video operations room to determine offsides and create limb-position 3D animations. This used to take around 70 seconds and now takes less than 30. Although this system is controversial, it is continuing to refine how sports refereeing grows.

7. Sustainability and Smart Energy Management

Large sports venues have always had a significant environmental impact and have come under increased scrutiny for it. In response, smart stadiums now aim to reduce this impact with real-time energy management systems. Using IoT and AI, smart sensors provide real-time data about energy use and waste. Smart lighting systems can automatically adjust to available natural light and the number of people in a room. Similarly, climate control systems use real-time data to maintain an optimal temperature depending on the number of people present, the conditions outside and the type of event.

8. Drone Cameras and Immersive Broadcasting Technology

An advantage of watching sports from home is that you can enjoy the broadcast experience, which includes commentary, multiple camera angles and instant replays. However, renovation of smart stadiums changes that advantage by providing the broadcast experience inside the stadium. Drone cameras were used inside the stadiums for the MLB All-Star game and some stadiums are equipping themselves with 360-degree cameras that will allow users to choose their viewing angle.

9. Blockchain Ticketing

The rise of blockchain technology offers a solution to what is perhaps the oldest problem in sports, ticketing fraud. Ticketing systems based on blockchain technology are the first to solve the problem of ticket fraud on a foundational level. Because tickets cannot be counterfeited, the resale of tickets occurs on an immutable ledger. Aside from fraud deterrence, blockchain technology paves the way for other engagements from fans, such as digital collectibles, loyalty rewards NFTs and tokens for access to exclusive experiences.

10. Improved Pricing and Smart Seating

Adaptive venues do not simply sell tickets; they sell each seat. Using algorithms, tickets are dynamically priced based on demand, opponent, weather, seat location and several other factors to maximise revenue. This practice also allows consumers the opportunity to purchase tickets at a lower price, so long as they are flexible with their seating.

The post 10 Amazing Features Of A Smart Stadium appeared first on TechRound.

Particular Audience is an AI-native retail media and personalisation company. The platform (DiscoveryOS) tackles the ‘endless aisle’ challenge by creating a unique storefront for every shopper – matching each visitor with the products they’re most likely to buy, at the moment they’re most likely to buy them.

How did you come up with the idea for the company?

I was running Asia-Pacific for a Google-backed personalisation startup, which recently sold to Publicis Group, and it was there that it occurred to me that legacy approaches to personalisation were super-manual and failed to drive improvements to the customer experiences.

As I started researching how Amazon succeeded in creating a different website for every customer, given their 300 million users and 300 million items, something clicked. Machine learning recommendation systems were foundational in enabling everything we now take for granted on the internet, from Facebook and Amazon to Netflix and Spotify, yet the majority of retailers were massively behind the curve, some still are.

We launched with what has now become the market-leading personalisation platform, powered by AI and machine learning. A year into operations, we noticed ‘Sponsored’ labels appearing across product listings on Amazon, and we were inspired. The amount of margin in selling a click is 4-10x the margin in shipping physical products, so we got to work in enabling this adtech functionality within our retail software-as-a-service.

Tell us about your core product or service

Retail Media is the topic du jour in most retail boardrooms today. The problem facing the market, however, is that nearly all software solutions in the market are strictly and simply advertising technology.

Retail media is about retail which provides real-time context of high intent consumers in transactable environments. This is immensely valuable to brands, and it is in stark contrast to ads on a news website, for example.

Most retailers face what we call a ‘multi-objective problem’ set. They may have suppliers wanting to fund promotional activity, yes, but they also have overstocks, private label items to push, spoiling and slow-moving inventory to shift, buyer-led trade agreements to honour, and visual merchandising imperatives to consider.

99% of retailers today face fragmentation both in their org chart, and especially in their technology vendors – one for search, one for personalisation, one for merchandising, one for pricing, one for retail media, etc.

Particular Audience brings this all under one decision engine, where different teams can collaborate within one unified customer experience technology. It might seem obvious, but retail is only now waking up to this. Our vertical integration of retail media with our own advanced AI-powered search and recommendation capabilities is what drastically improves ad performance vs siloed platforms. It is unique and it is immensely powerful for maturing retail media networks that face tradeoffs in ad density.

What most excites you about the retail media industry?

We are still so early in its development. Only a tiny minority of retailers, albeit some of the largest, have launched retail media networks so far, and massive agency budgets are only just opening up to retail media.

Retail media teams, often with advertising backgrounds, are quickly learning about the nuances of retail and are on the same journey that e-commerce and conversion rate optimisation teams have been on for the last decade or so. This knowledge curve is unlocking tremendous value once legacy technology is replaced.

What has been the biggest challenge you’ve had to overcome along the way?

I think it’s really hard to run a startup without spending 110% of your waking hours thinking and acting on the problem space you have chosen to build for, and to approach it all with first-principles thinking.

Everyone has an opinion, often informed by anecdotal experience in adjacent and therefore distinct situations. Talk is cheap. I would say my biggest challenge has been maintaining laser-focused conviction throughout the marathon. There are always bumps in the road, and you get pulled in all sorts of directions by the market, cheap advice and shiny-object syndrome.

What is your number-one piece of advice to aspiring entrepreneurs?

Resilience is what separates winners and losers. Market-leading enterprise software platforms aren’t built overnight; no startup is born enterprise-ready. Success is a product of years of client feedback and feature requests to refine a defensible compound value proposition. The trick is staying alive. Be conservative with money but take measured risks. Learn to price risk.

What can we hope to see from Particular Audience in the future?

AI-as-a-channel is an exciting opportunity we are working with a select few retailers on using what I think is the most exciting technology in the market right now: Model Context Protocol. A big question is, what happens to sponsored listings in retail media when consumers use AI assistants to shop on their behalf? Our answer is that retail media needs to be further integrated with organic discovery on websites to succeed here, and Particular Audience is uniquely positioned to make it happen for retailers.

The post Meet James Taylor, Founder and CEO at Particular Audience, The Company Bringing Amazon-Style Hyper-Personalisation To Every Other Retailer appeared first on TechRound.