Author: Eternity Lab

Panel members on the AI as a Competitive Advantage session held at the AI Summit in London this week discussed the reality businesses face when trying to move artificial intelligence projects into production.

Data presented in the Summit’s AI at Scale stream, suggests that 80% of proof-of-concept AI projects fail to move into production. While the panel discussion did not focus heavily on moving beyond AI proof-of-concept projects, as that topic was discussed in a previous session, two members of the panel did raise the issue of how AI aligns with enterprise IT.

This can be a challenge for moving beyond proof of concepts, especially when the new technology being piloted needs to integrate with existing IT infrastructure and enterprise datasets.

Ravi Rabheru, head of AI centre of excellence at Intel for EMEA, noted that the big challenge businesses face is around technical debt.

Dara Sosulski, head of AI and model management for HSBC, added: “The bigger the company, the more the technical debt and the more the complexity.”

It is an area of concern both in large enterprises and in government, which has an agenda to push AI-enablement across the public sector. The Public Account Committee’s (PAC’s) Use of AI in government report from March 2025 noted that AI relies on high-quality data to learn. However, the committee was told by the Department for Science, Innovation and Technology (DSIT) that government data is often of poor quality and locked away in out-of-date legacy IT systems.

Sosulski noted that IT leaders need to assess whether their data infrastructure is right for AI applications to prioritise and understand what is achievable: “Infrastructure is the thing that unlocks the keys of the Kingdom, in a way. You then have something that is a backbone and it’s modular and interoperable. [With such IT infrastructure], you can access applications from other places and you can connect to other things.” 

However, she acknowledged that it may not be possible to provide a date as to when all the components needed for AI will be in place in some organisations.

Build or buy?

But the industry is keen to promote the value of ready-made AI capabilities. Sosulski believes that the build-versus-buy question essentially comes down to the use cases the business wants to tackle, saying: “I think all enterprises have now adopted a very similar set of tools to solve problems like software development, drafting emails, translating a document and document Q&A.”

Given that there are products that cater for such use cases, she added: “There are some use cases that are so generalist, we would consider them core capabilities. Those are ones that we consider buying as an enterprise wide solution that is tested and integrates well with our other IT infrastructure. Everybody realises you don’t solve them internally at great expense.” 

While business and leaders tackle technical debt and balance when to build and when to buy AI functionality, they also need to keep abreast of the latest developments.

While the whole tech industry appears to be steamrolling agentic AI and artificial general intelligence (AGI), Sosulski recommended that technology decision-makers look at what developments are relevant to the business.

Sosulski felt that there’s less of a need to keep up with the latest AI foundation model. “Despite everything changing constantly, a lot of these models wind up being a much of a muchness for what you want to do,” she said. “We don’t need new foundation models every six months. HSBC and most companies are that way and so, at some point, you just get familiar with the ins and outs of what the models can and can’t do.” 

With a selection of some open source and proprietary models, Sosulski urged delegates to focus on assessing which models work best for their use cases. These can then be piloted in proof-of-concept projects to prove they work. She also recommended putting in a place a control framework and IT infrastructure that enables retraining and iterating quickly.

With such a setup, she said: “You can keep moving things out into production without having to overhaul everything every six or 12 months.”

Source

Meta launched the Meta AI app in late April to take on ChatGPT and other chatbots. Unlike rival apps, Meta AI comes with social features that nobody asked for. But Meta’s desire for Meta AI users to share their chats with others via a social feed isn’t surprising. Social media is how Meta makes its money. All of its apps are social apps. Also, bringing a social element to an AI chatbot experience could always work in Meta’s favor.

However, that’s hardly the case right now. Meta AI has gone viral this week for a huge issue. Rather than discussing a unique Meta AI feature that makes the chatbot a must-have AI product, people are talking about the wildly inappropriate chats that take place on the platform, which some users are sharing online by mistake for others to see.

Sharing AI chats is optional, but it looks like plenty of users don’t realize what they’re doing, or they don’t care. Whatever the case, the Meta AI chats that appeared on social media are deeply disturbing. They show what can go wrong if an AI firm working on frontier AI experiences doesn’t handle user privacy correctly. Meta could do a better job informing users that the “Share” button will move the Meta AI chat to the Discover feed.

According to TechCrunch, around 6.5 million people installed the standalone Meta AI app. The figures come from Appfigures, not Meta. That’s hardly the user base that a company like Meta can brag about. Meta AI is a standalone app. It wasn’t embedded in a more popular app like Instagram or WhatsApp, so it’s up to users to install it.

Tech. Entertainment. Science. Your inbox.

Sign up for the most interesting tech & entertainment news out there.

By signing up, I agree to the Terms of Use and have reviewed the Privacy Notice.

Before rolling the app out, Meta largely focused on forcing Meta AI experiences into all its social apps, including WhatsApp, Messenger, Instagram, and Facebook. That’s why Meta can say Meta AI has 1 billion monthly users.

The standalone Meta AI app has yet to achieve such reach. But even so, 6.5 million isn’t a small number. It shows that some people are genuinely interested in the Meta AI chatbot experience. However, not all of them know how to protect their privacy.

I haven’t tried Meta AI, nor am I likely to get on the app anytime soon. But privacy is one of my main concerns when it comes to AI products. Meta could do a better job here. While I haven’t been exposed to private Meta AI chats that were shared online by users who don’t know (or care) about how the social aspect works, there are plenty of examples.

Here’s a take from TechCrunch:

Flatulence-related inquiries are the least of Meta’s problems. On the Meta AI app, I have seen people ask for help with tax evasion, if their family members would be arrested for their proximity to white-collar crimes, or how to write a character reference letter for an employee facing legal troubles, with that person’s first and last name included. Others, like security expert Rachel Tobac, found examples of people’s home addresses and sensitive court details, among other private information.

It keeps going, too. Here’s what Gizmodo found in the Discover feed, which is where the Meta AI chats go if you don’t know what you’re doing and press the Share button:

In my exploration of the app, I found seemingly confidential prompts addressing doubts/issues with significant others, including one woman questioning whether her male partner is truly a feminist. I also uncovered a self-identified 66-year-old man asking where he can find women who are interested in “older men,” and just a few hours later, inquiring about transgender women in Thailand.

Andreessen Horowitz partner Justine Moore posted screenshots of Meta AI chats in the Discover feed, summarizing some of what she saw in an hour of browsing:

• Medical and tax records
• Private details on court cases
• Draft apology letters for crimes
• Home addresses
• Confessions of affairs…and much more!

None of these topics should be broached in your conversations with any AI model, whether it’s Meta AI, ChatGPT, or any of the other countless startups.

What you can do

If you or someone you love is using Meta AI, you should ensure the privacy settings are set correctly. Gizmodo, which hilariously advises users to get their parents off of Meta AI, lists the steps needed to prevent Meta AI chats from making it to the Discover feed:

1. Tap your profile icon at the top right.
2. Tap “Data & Privacy” under “App settings.”
3. Tap “Manage your information.”
4. Then, tap “Make all your prompts visible to only you.”
5. If you’ve already posted publicly and want to remove those posts, you can also tap “Delete all prompts.”

Also, don’t tap the Share button if you want to keep an AI conversation private. As TechCrunch and Justine Moore point out, the Meta AI chats are not public by default. But some people press the Share button in chats, unaware they’re sharing them with everyone else on the platform.

I’ll also remind you that Meta will use all your posts shared on its social networks to train AI. You might want to opt out of that if you haven’t done so already. And if you don’t want Meta AI to use any of that public information for more personalized responses, you’ll want to opt out of that too.

Finally, remember that it’s not just older, less tech-savvy people using Meta AI in ways that might be inappropriate. You’ll want to check on your teens as well and see what sort of chats they might have with Meta AI.

Source

Two years after the merger of the two leading UK connectivity companies completed, VodafoneThree is now set to challenge the mobile arena and the UK’s broadband market by aiming to cover all homes with fibre or fibre-like speeds, boosting this ambition with a new partnership with Community Fibre.

Vodafone and Three first announced plans to merge in June 2023 and the new combined business will operate under the VodafoneThree brand, with 51% owned by Vodafone Group and 49% by CK Hutchison Group Telecom (CKHGT) Holdings. Vodafone will fully consolidate VodafoneThree in its financial results and the company’s CEO will be Max Taylor, who currently leads Vodafone UK. Three UK’s Darren Purkis has been appointed chief financial officer (CFO).

Backed by an £11bn infrastructure investment – one of the key guarantees by the two companies to gain closure of the deal from UK regulatory authorities – to connect all four nations and every community, VodafoneThree will be the biggest mobile network operator in the UK, with 27 million customers.

In addition, it said that it is the only UK operator with a quarter-by-quarter, year-by-year, guaranteed plan to reach 99.95% 5G Standalone (5G SA) population coverage by 2034. The 5G SA network build-out plan is front-loaded so that it will hit 90% population coverage from a current baseline of 47% by the end of the third year.

VodafoneThree assured that up to 50 million people will have access to its fastest 5G speeds in just one year. Explaining why it is emphasising 5G SA, the company said only such networks will have the capacity and speed to manage the vast amounts of data that applications such as AI will require, meaning that the target is to build the first nationwide, AI-ready network.

The new merged company will operate a multi-brand mobile strategy in consumer with Vodafone, Three, VOXI, Smarty and Talkmobile remaining, while Vodafone will be the only brand for business customers, with one team able to tailor systems to a customer’s needs, with the ambition to become the UK’s biggest converged network for business.

Over the course of the next 12 months, VodafoneThree will bring Three’s fixed wireless access services together with Vodafone’s full-fibre offer into one home broadband portfolio under the Vodafone brand. There will only be one converged brand for both businesses and consumers.

Through the sharing of combined spectrum, VodafoneThree said that within two weeks, seven million Three and Smarty customers will receive a 4G boost, with an improvement in 4G data speed of up to 20%. Within a few months, all legacy Vodafone and Three mobile customers will gain access to roam on each other’s networks at no extra cost.

This is intended to remove a total of 16,500 sq/km of “not spots” – said to be equivalent to 10x the size of London – by the end of the 2025, with the first sites already having been turned on. Taking advantage of Vodafone’s recent partnership with AST Space Mobile, VodafoneThree said that it will launch beta trials on a “first-of-its-kind”, space-based satellite mobile network by June 2026. The space communications platform will complement the existing network build, eliminating coverage gaps in places that otherwise couldn’t be reached.

While much of the merger focus has been on mobile, VodafoneThree said that it will challenge the broadband market, bringing better value and choice through existing and new partnerships. The new company boasts more than million customers and can currently bring fibre or fibre-like speeds of up to 2.2Gbps to 22.5 million homes and businesses.

The merger has attracted support from across the communications industry and from the world of politics. Commenting on what it could achieve, chancellor of the exchequer Rachel Reeves, said: “I’m delighted that that this huge investment is being made in mobile phone network infrastructure, better connecting people with families, loved ones and work by providing stronger, more widespread 5G coverage.

“Our Spending Review is all about growing our economy and attracting private investment to deliver on our Plan for Change, and it’s fantastic that VodafoneThree have chosen to join us on our mission of national renewal.”

Source

With WWDC 2025 behind us, developers are now testing the first iOS 26 beta. Meanwhile, many iPhone owners are wondering when they will have a chance to preview the upcoming features. Here’s everything you need to know about the availability of the iOS 26 public beta.

Did Apple announce when users will be able to download the iOS 26 public beta?

Yes, Apple’s CEO Tim Cook announced during the WWDC 2025 keynote that the public beta of iOS 26 will be available in July. Apple usually releases a few betas exclusively for developers before opening up beta testing to the general public.

When should you expect the first public beta?

We expect to see iOS 26 public beta 1 between the second and third week of July. Here’s when Apple released the first public beta of previous iOS updates in the past few years:

• iOS 15: June 30, 2021 
• iOS 16: July 11, 2022 
• iOS 17: July 12, 2023
• iOS 18: July 15, 2024 

We wouldn’t be surprised if the public beta of iOS 26 was released around July 14 or July 15.

Tech. Entertainment. Science. Your inbox.

Sign up for the most interesting tech & entertainment news out there.

By signing up, I agree to the Terms of Use and have reviewed the Privacy Notice.

How can you enroll in the beta program?

If you want to be one of the first to try the public beta, you need to enroll your Apple Account in the Apple Beta Software Program. Here are the steps to follow:

• First, make sure you’re enrolled in the Apple Beta Software Program
• On your iPhone, open the Settings app
• Tap on General, and then Software Update
• Tap on Beta Updates and enable iOS 26 Public Beta

These are some of the features you should expect to see on iOS 26 public beta 1

Image source: José Adorno for bGR

Once the public beta is available, you’ll be able to try out the following features:

• All-new design: Apple offers a brand new design with the Liquid Glass UI. It’s a drastic change from the previous design language of iOS 18.
• Messages app improvements: Users can screen messages from unknown senders, create polls, and generate unique backgrounds with Image Playground.
• Apple Music: Alongside a revamped app, users can pin favorite songs on top of the library, in addition to several other changes.
• ChatGPT improvements: ChatGPT features onscreen awareness for those who use it to power Apple Intelligence.
• Genmoji and Image Playground updates: Apple says iOS 26 improves these OG Apple Intelligence features as it’s now possible to mix emoji, Genmoji, and descriptions together to create something new.

Supported devices

Notch on the iPhone 16e. Image source: Christian de Looper for BGR

These are all of the devices that will support iOS 26:

• iPhone 11
• iPhone 11 Pro and 11 Pro Max
• iPhone SE (2nd gen)
• iPhone 12 mini and iPhone 12
• iPhone 12 Pro and iPhone 12 Pro Max
• iPhone 13 mini and iPhone 13
• iPhone 13 Pro and iPhone 13 Pro Max
• iPhone SE (3rd gen)
• iPhone 14 and iPhone 14 Plus
• iPhone 14 Pro and iPhone 14 Pro Max
• iPhone 15 and iPhone 15 Plus
• iPhone 15 Pro and iPhone 15 Pro Max
• iPhone 16 and iPhone 16 Plus
• iPhone 16 Pro and iPhone 16 Pro Max
• iPhone 16e

Below, you can learn more about some of the other new features coming in iOS 26.

Source

If you visited adult sites on your Android device via incognito browsing sessions or a VPN while also logged into Facebook and Instagram, Meta might know about it. Not only that, Meta might know what you watched, what you clicked, and what you bought.

Regardless of what you were doing on your device, if you did it on an Android phone while logged into Meta’s social networks, Meta was likely tracking it.

What will Meta do with all that extra data it collects about you? It’ll use the data to serve personalized ads that match your most recent interests.

The method Meta used is incredibly abusive, as the company did it without obtaining consent from the user. Instead, Meta used Android browsers to link browsing history to Facebook and Instagram profiles. Even Google, which has its own history of tracking users without consent, called the procedure a “blatant violation” of its security principles.

Tech. Entertainment. Science. Your inbox.

Sign up for the most interesting tech & entertainment news out there.

By signing up, I agree to the Terms of Use and have reviewed the Privacy Notice.

Meta stopped the tracking this week, but even if it were to continue, Google is already patching Chrome. I wouldn’t be surprised to see regulators investigate Meta’s behavior next, especially in jurisdictions like the EU, which have stronger privacy protections for users.

Meta’s abuse was discovered accidentally by professor Günes Acar from Radboud University in the Netherlands (here’s his research). He observed an unexpected connection between a page containing trackers, including Facebook, and his computer.

After a while, Acar realized that Facebook had hidden its tracking method so that it would be more difficult to spot. Meta found a way to link the Meta Pixel tracker that’s present on around 20% of the most popular websites, including those featuring adult content, to the Facebook and Instagram apps installed on an Android phone.

Normally, the Meta Pixel tracker would send information to Meta. But Meta found a way to connect Meta Pixel to the cell phone app. This linked the web traffic it recorded to the user’s identity, as registered in Facebook and Instagram.

This allowed Meta to collect browsing information from someone’s session without asking for consent. The method also bypassed the privacy protections you might have enabled on your Android phone, like incognito mode or VPN. These are means to anonymize internet traffic and prevent anyone from tracking you. We’ve known for quite a while that incognito mode isn’t truly anonymous, but VPN protections can offer stronger privacy.

The method Meta devised ignores these protections, allowing Meta to capture all your actions on a web page and link it to your Facebook and Instagram profiles. Meta would know what products you looked at and what you bought. In turn, this will let it serve more personalized ads based on these recent activities.

Google had a stark response when asked about the new security research, providing El País with the following statement:

The developers mentioned in this report are unintentionally using functions present in many iOS and Android browsers, which blatantly violate our security and privacy principles. We have already implemented changes to mitigate these invasive techniques, launched our own investigation, and are in direct contact with the parties involved.

These remarks are especially notable considering where they’re coming from. Google recently settled a user tracking lawsuit for nearly $1.4 billion. It knows a thing or two about tracking users without consent.

Meta didn’t exactly claim responsibility. It’s referring to the incident as “an issue” that it’s working with Google to resolve. Here’s Meta’s statement on the matter:

We are speaking with Google to clarify a potential misunderstanding about how its policies are applied. As soon as we learned of the concern, we decided to pause the feature while we work with Google to resolve the issue.

Meta didn’t explain why it’s been tracking web browsing without consent. One possibility is that Google’s initiative to limit third-party tracking in browsers, which has been abandoned, motivated Meta to find other ways to track users.

Even if Meta decides not to discontinue this sort of tracking, Google is patching Chrome. Similarly, Mozilla is developing a fix to prevent such tracking. Separately, DuckDuckGo said Meta’s trackers did not affect its users.

But Meta isn’t alone in using such shady tactics. Russian company Yandex employed similar tools to track users online. Meta has only been taking this approach since September 2024, while Yandex started doing it in 2017.

Meta did protest Apple vehemently when the latter introduced strong privacy protections on the iPhone. Those features, which help users prevent tracking on apps like Facebook and Instagram, cost Meta billions in lost advertising as soon as Apple introduced them.

More recently, Meta asked Apple for unprecedented access to iPhone data under the Digital Markets Act (DMA) law in Europe. Apple has already raised alarms about Meta’s intentions:

As we strive to comply with the DMA, we carefully review each interoperability request we receive. As an example of our concerns, Meta has made 15 requests (and counting) for potentially far-reaching access to Apple’s technology stack that, if granted as sought, would reduce the protections around personal data that our users have come to expect from their devices.

If Apple were to have to grant all of these requests, Facebook, Instagram, and WhatsApp could enable Meta to read on a user’s device all of their messages and emails, see every phone call they make or receive, track every app that they use, scan all of their photos, look at their files and calendar events, log all of their passwords, and more. This is data that Apple itself has chosen not to access to provide the strongest possible protection to users.

You’ll find El Pais’ coverage at this link. Also, this website explains in great detail how Meta might have tracked your web browsing on Android.

Source

Artificial intelligence (AI) factories are producing intelligence at unprecedented scale, showing massive potential to fuel economic growth and innovation.

To accelerate taking advantage of AI opportunities in their sector, operators Orange, Fastweb, Swisscom, Telefónica and Telenor have announced a collaboration with Nvidia to develop and expand sovereign AI factories and edge infrastructure across Europe.

The telcos believe that they are uniquely positioned to deliver AI services due to their extensive infrastructure and geographic reach. The ultimate aim of the partnership is to equip European enterprises across industries with secure, accelerated computing infrastructure to train and deploy customised AI models and agentic AI services working in a sovereign AI infrastructure.

The collaboration will see Nvidia’s full-stack technologies implemented to allow telcos to go beyond traditional connectivity services to AI as a service, driving generative and agentic AI adoption. This is also seen as empowering nations to build localised models with secure sovereign AI infrastructure.

Looking at individual companies’ deployments, the enterprise division of the Orange Group, Orange Business, has joined the Nvidia Cloud Partner programme to accelerate the development of enterprise-grade agentic AI, including its Live Intelligence platform, which is designed to enable companies to securely deploy generative AI at scale. Those AI solutions use the Orange Business Cloud Avenue platform, built on Nvidia infrastructure.

This is seen as allowing enterprises across Europe to access AI infrastructure, train custom AI models and deploy secure generative AI applications through the Orange Business’ Live Intelligence platform – as well as offer the possibility of transforming employee, customer and operational experiences.

Orange has also applied AI to its own operations, with 73,000 employees now regularly using the system to streamline tasks, develop software, automate support procedures and enhance decision-making processes. Orange’s sovereign AI infrastructure supports employees in France, as well as across Europe and Africa, and handling more than 30,000 requests daily.

Building on its creation of its host country’s first sovereign AI infrastructure, Norway’s Telenor is to expand capacity to meet rising demand from both internal teams and external customers. With plans to add an AI datacentre that will run entirely on renewable energy and contribute surplus energy back to the grid, the company believes it is helping Norway to advance its mission of secure, sovereign and sustainable AI development.

Telenor believes that since launch, its AI infrastructure has helped drive AI adoption across Norway by powering digital services for the public sector, industrial automation and local language models.

This includes hosting BabelSpeak, an AI-driven translation tool developed by Capgemini that offers near-real-time, voice-to-voice multilingual translation capabilities in nearly 100 languages and is now being piloted by Norwegian Red Cross.

Capgemini has been a key partner for Telenor to build Norway’s first sovereign and secure AI cloud service in collaboration with Nvidia.

In addition, Telenor is integrating Nvidia AI Enterprise software to accelerate enterprise adoption and deployment of generative and agentic AI applications, as well as its own internal innovation efforts, such as in network automation.

Swisscom has launched the AI Work Hub and Model catalogue, enabling enterprises across Switzerland to build complex AI projects, customise models and deploy agentic AI at scale and speed.

The operator recently announced GenAI Studio – a new service built on its Swiss AI Platform that allows enterprises to develop and run AI agents. The enterprise AI services are hosted on Swisscom’s sovereign AI factory, built on Nvidia DGX SuperPOD. This allows for scaling of capacity to serve Switzerland’s rapidly growing demand for AI services and inference, expanding the company’s revenue opportunities.

Source

Just days after launching a wholesale multi-gig broadband product designed to enable internet service providers (ISPs) to offer business customers connectivity with symmetrical speeds of 5.5Gbps, CityFibre has announced that it has tripled the availability of its dedicated, enterprise-grade Ethernet services to serve more than 260,000 UK businesses.

The roll-out incorporates a dedicated business ethernet platform and is supplied by CityFibre’s engineering partner Calix. It is designed specifically to support enterprise-grade service guarantees and brings CityFibre’s full FTTP network online for its Business ISPs.

The increase not only expands the total number of businesses available to partners – rising from 80,000 – but is also said to increases the density of serviceable businesses in its network footprint, making it easier and more efficient for partners to run targeted sales and marketing campaigns. The provider said that it was also committed to delivering “the best economics”, claiming that its Ethernet prices were up to 20% cheaper than its major competitors.

CityFibre said that its investment in its Ethernet services has been underpinned by its nationwide 10Gbps XGS passive optical network (XGS-PON) technology upgrade being rolled out across its fibre-to-the-premises (FTTP) network. This began in 2022 and is already implemented across 85% of CityFibre’s network and is due for completion later this summer. The 5.5Gbps wholesale product will be made available across CityFibre’s nationwide full-fibre network, with faster multi-gig services due to launch in 2026.

“[This] announcement is a reminder that CityFibre is more than just a broadband network. We’re building a digital infrastructure platform that serves the needs of all sectors of the market, from consumer services right through to enterprise-level connectivity,” said Andy Nash, CityFibre director for business, government and mobile.

“Our partners have called for a larger footprint and that is exactly what we are giving them. Tripling our business footprint transforms their addressable market and is a catalyst for growth and shared success. We look forward to our proposition of market leading products, service and economics benefitting even more businesses in the years ahead.”

From a strategic basis, CityFibre said that over the past year, it has worked closely with its partners to speed-up the installation of new business Ethernet connections by up to 40% across its most complex Ethernet installations with permission to work agreed.

CityFibre also revealed that it was introducing a delivery assurance guarantee, providing partners with the confidence of enhanced compensation for any unexpected delays to their installation date.

“As UK businesses increasingly rely on technology to operate, reliable, secure, fast internet access is becoming a strategic asset,” said Nathan Marke, chief strategy officer at CityFibre partner, Giacom.

“The delivery assurance guarantee is a great example of the CityFibre challenger DNA and difference. They have listened to the market and focused not just on great technology, but also predictable installations and great service. This brings invaluable peace of mind for our community of more than 6,000 MSPs and resellers as they deliver critical customer projects.”

Nick Shraga, chief operating officer at Southern Communications, added: “Our customers are growing and adapting fast, powered by outstanding connectivity alongside brilliant customer service. So, CityFibre’s Ethernet expansion, combined with its commitment to competitive pricing and reliable installations, is great news and promises to be a boost for UK businesses.”

Source

Technology leaders are baffled by a “cacophony” of “buzzwords, hype and confusion” over the benefits of artificial intelligence (AI), according to the founder and CEO of technology company Pegasystems.

Alan Trefler, who is known for his prowess at chess and ping pong, as well as running a $1.5bn turnover tech company, spends much of his time meeting clients, CIOs and business leaders.

“I think CIOs are struggling to understand all of the buzzwords, hype and confusion that exists,” he said.

“The words AI and agentic are being thrown around in this great cacophony and they don’t know what it means. I hear that constantly.”

CIOs are under pressure from their CEOs, who are convinced AI will offer something valuable.

“CIOs are really hungry for pragmatic and practical solutions, and in the absence of those, many of them are doing a lot of experimentation,” said Trefler.

Companies are looking at large language models to summarise documents, or to help stimulate ideas for knowledge workers, or generate first drafts of reports – all of which will save time and make people more productive.

The risk of rogue AI

But Trefler said companies are wary of letting AI loose on critical business applications, because it’s just too unpredictable and prone to hallucinations.

“There is a lot of fear over handing things over to something that no one understands exactly how it works, and that is the absolute state of play when it comes to general AI models,” he said.

Trefler is scathing about big tech companies that are pushing AI agents and large language models for business-critical applications. “I think they have taken an expedient but short-sighted path,” he said.

“I believe the idea that you will turn over critical business operations to an agent, when those operations have to be predictable, reliable, precise and fair to clients … is something that is full of issues, not just in the short term, but structurally.”

One of the problems is that generative AI models are extraordinarily sensitive to the data they are trained on and the construction of the prompts used to instruct them. A slight change in a prompt or in the training data can lead to a very different outcome.

For example, a business banking application might learn its customer is a bit richer or a bit poorer than expected.

“You could easily imagine the prompt deciding to change the interest rate charged, whether that was what the institution wanted or whether it would be legal according to the various regulations that lenders must comply with,” said Trefler.

Thinking in advance is better than thinking on your feet

Trefler said Pega has taken a different approach to some other technology suppliers in the way it adds AI into business applications.

Rather than using AI agents to solve problems in real time, AI agents do their thinking in advance.

Business experts can use them to help them co-design business processes to perform anything from assessing a loan application, giving an offer to a valued customer, or sending out an invoice.

Companies can still deploy AI chatbots and bots capable of answering queries on the phone. Their job is not to work out the solution from scratch for every enquiry, but to decide which is the right pre-written process to follow.

As Trefler put it, design agents can create “dozens and dozens” of workflows to handle all the actions a company needs to take care of its customers.

“You just use the natural language model for semantics to be able to handle the miracle of getting the language right, but tie that language to workflows, so that you have reliable, predictable, regulatory-approved ways to execute,” he said.

The right AI for the job

Large language models (LLMs) are not always the right solution. Trefler demonstrated how ChatGPT 4.0 tried and failed to solve a chess puzzle. The LLM repeatedly suggested impossible or illegal moves, despite Trefler’s corrections. On the other hand, another AI tool, Stockfish, a dedicated chess engine, solved the problem instantly.

The other drawback with LLMs is that they consume vast amounts of energy. That means if AI agents are reasoning during “run time”, they are going to consume hundreds of times more electricity than an AI agent that simply selects from pre-determined workflows, said Trefler.

“ChatGPT is inherently, enormously consumptive … as it’s answering your question, its firing literally hundreds of millions to trillions of nodes,” he said. “All of that takes [large quantities of] electricity.”

Using an employee pay claim as an example, Trefler said a better alternative is to generate, say, 30 alternative workflows to cover the major variations found in a pay claim.

That gives you “real specificity and real efficiency”, he said. “And it’s a very different approach to turning a process over to a machine with a prompt and letting the machine reason it through every single time.”

“If you go down the philosophy of using a graphics processing unit [GPU] to do the creation of a workflow and a workflow engine to execute the workflow, the workflow engine takes a 200th of the electricity because there is no reasoning,” said Trefler.

He is clear that the growing use of AI will have a profound effect on the jobs market, and that whole categories of jobs will disappear.

The need for translators, for example, is likely to dry up by 2027 as AI systems become better at translating spoken and written language. Google’s real-time translator is already “frighteningly good” and improving.

Pega now plans to work more closely with its network of system integrators, including Accenture and Cognizant to deliver AI services to businesses.

Major expansion

An initiative launched last week will allow system integrators to incorporate their own best practices and tools into Pega’s rapid workflow development tools. The move will mean Pega’s technology reaches a wider range of businesses.

Under the programme, known as Powered by Pega Blueprint, system integrators will be able to deploy customised versions of Blueprint.

They can use the tool to reverse-engineer ageing applications and replace them with modern AI workflows that can run on Pega’s cloud-based platform.

“The idea is that we are looking to make this Blueprint Agent design approach available not just through us, but through a bunch of major partners supplemented with their own intellectual property,” said Trefler.

That represents a major expansion for Pega, which has largely concentrated on supplying technology to several hundred clients, representing the top Fortune 500 companies.

“We have never done something like this before, and I think that is going to lead to a massive shift in how this technology can go out to market,” he added.

Source

Apple HQ has seemingly sprung a leak in the days ahead of its WWDC 2025 keynote, as new details about the company’s upcoming software updates continue to find their way online. After spoiling a few surprises pertaining to iOS 26 on Tuesday, 9to5Mac returned today with updates about new features likely coming to Apple’s AirPods this fall.

According to 9to5Mac, there are several new features currently in the works for Apple’s line of wireless earbuds. The expectation is that these new features will be included in the iOS 26 and macOS 26 software updates, but the site also warns that Apple may “choose to delay or tweak” these features, so there’s no guarantee we’ll see them this fall.

With that out of the way, let’s dig into the rumored AirPods upgrades.

New AirPods features set for iOS 26

AirPods Pro 2 with USB-C port. Image source: Christian de Looper for BGR

Of the rumored features, the most intriguing is camera control. 9to5Mac says AirPods users will be able to press the AirPods stem in order to snap a photo with a connected iPhone. This exact feature was available on Apple’s EarBuds, but has yet to appear on AirPods.

Tech. Entertainment. Science. Your inbox.

Sign up for the most interesting tech & entertainment news out there.

By signing up, I agree to the Terms of Use and have reviewed the Privacy Notice.

The report also notes that AirPods will soon support new head gestures. Apple introduced head gestures on AirPods Pro 2 and AirPods 4 in 2024, allowing users to move their heads to answer or decline phone calls as well as interact with messages and notifications. On iOS 26, Apple will reportedly add even more gestures, potentially including one that would give users the ability to end or extend a Conversation Awareness volume adjustment.

Other new features coming to Apple’s AirPods in iOS 26 include sleep detection to automatically pause content when the user dozes off, a “studio quality” mic mode similar to the iPhone’s Audio Mix feature, and an improved pairing experience.

Tune in to WWDC 2025

WWDC 2025 begins on June 9. Image source: Apple Inc.

As for which AirPods will get these features, we’ll have to wait until Monday to find out. Apple’s WWDC 2025 keynote is scheduled to begin on June 9 at 10:00 a.m. PT / 1:00 p.m. ET. Even if these new features aren’t announced during the keynote, we always find a wide variety of new features that Apple didn’t have time to discuss once the dust settles after the event.

We’ll be keeping our eyes peeled for these AirPods upgrades and more on Monday.

Source

Apple fans around the world are probably anxiously waiting for next week’s main keynote, which will open this year’s WWDC. The event is where Apple unveils its latest operating systems each year, making them available in beta to testers eager to try the new features right after the show.

This year’s OS reveals are even more exciting, since we’re not just getting new features for iPhone, iPad, Mac, Apple Watch, Apple TV, AirPods, and Vision Pro. Apple is giving all its operating systems a big redesign, one that has leaked several times over the past few months.

A name change is also coming, with Apple replacing the expected iOS 19 label with a new iOS 26 moniker meant to designate the year of release. The official iOS 26 launch will happen in September or late 2025, making it the newest iPhone OS for most of 2026.

This new naming policy will apply across the board. All of Apple’s operating systems will follow the same numbering scheme.

Tech. Entertainment. Science. Your inbox.

Sign up for the most interesting tech & entertainment news out there.

By signing up, I agree to the Terms of Use and have reviewed the Privacy Notice.

With all that in mind, iOS 26 is easily the most anticipated OS coming out of WWDC. It’s the operating system people will want to try right away on their smartphones, to see how the new design works and test whatever iOS 26 features are available to early adopters on day one.

Unfortunately, some older iPhones still in use won’t be able to run iOS 26.

Apple has an impressive track record when it comes to supporting older iPhones. The latest iOS features might not show up on phones that launched five years ago, but they’ll still get most of the iOS functionality and the new design.

Apple Intelligence will be one of the iOS 26 features limited to newer models, but that’s due to specific hardware requirements for AI.

Otherwise, the visionOS-inspired design in iOS 26 will be available on older devices. The same goes for new features built into iOS 26 apps.

The first iOS 26 beta won’t include all the new features Apple planned for this release. Only the ones ready for testing will be available to users.

Also, just a reminder that Apple gives developers first access to the beta. They get it at WWDC, and the public beta usually follows about a month later.

As for supported hardware, iOS 26 will work on nearly all iPhones that can run iOS 18. A report from a trusted source mentioned a few months ago that only the 2018 iPhones wouldn’t be upgraded to iOS 26 (still called iOS 19 at the time). Apple will confirm the list next week, but that same group of iPhones has resurfaced online.

The rumor is likely accurate. These iPhone models should support iOS 26 starting with the first beta release next week:

• iPhone 11, 11 Pro, and 11 Pro Max
• iPhone 12, 12 mini, 12 Pro, and 12 Pro Max
• iPhone 13, 13 mini, 13 Pro, and 13 Pro Max
• iPhone 14, 14 Plus, 14 Pro, and 14 Pro Max
• iPhone 15, 15 Plus, 15 Pro, and 15 Pro Max
• iPhone 16e, 16, 16 Plus, 16 Pro, and 16 Pro Max
• iPhone SE (2nd gen and 3rd gen)

If you’re currently in the iOS 18 beta program, you’ll move to iOS 26 once Apple rolls out the first beta.

Source