Author: Eternity Lab

One of the most common complaints among iPhone users is lens flare when taking photos. When the image features a bright light source like the Sun in the background, it’s likely that users will see some lens flare in their pictures.

Last year, leaker yeux1122 mentioned that Apple was testing a new camera lens coating to prevent iPhone lens flare. However, when the iPhone 16 Pro became available, we realized the issue wasn’t fixed, and Apple hadn’t included this miraculous fix.

Still, Patently Apple might offer some hope, as the US Patent Office recently published a patent application for Apple’s potential fix for its longtime lens flare problem.

According to the patent, Apple may use optically absorptive gratings in its camera modules. These gratings are ultra-thin layers with tiny shapes that absorb light.

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 grating blocks the “bad” light, so your iPhone only captures the “good” light. This could allow you to capture a beautiful picture of the Moon without any lens flare. As the patent explains, the tiny shapes can be arranged in straight lines or in a zigzag pattern, all built on top of a thin base.

This could make the next iPhone camera smarter and sharper, with photos looking clearer and better even in tricky lighting.

However, it’s still unclear when this feature will arrive. So far, the camera upgrades expected for the iPhone 17 Pro focus on resolution and functions, but not on this specific fix.

For example, BGR reported that Apple is planning a 48MP telephoto camera with a smaller zoom in 48MP but a 7x 12MP zoom. There are also rumors that Apple will allow users to record in 8K for the first time, and the front-facing camera is expected to get a 24MP lens.

Source

UK based legal professor Or Brook has filed a class action against Google worth approximately £5bn in the UK Competition Appeal Tribunal (CAT). The class action, brought on behalf of hundreds of thousands of UK-based organisations that used Google’s search advertising services, accuses Google of abusing its near-total dominance in the general search market to drive up prices.

This latest class action follows on from one filed by Nikki Stopford, co-founder of Consumer Voice, and legal firm Hausfeld & Co LLP, and appears to focus on the Google’s anti-competitiveness.

Stopford’s case looks at the cost to consumers due to increased advertising costs businesses that use Google Search pay as a result of anti-competitive practices. In November last year, Google’s attempt to throw out Stopford’s case was dismissed, paving the way for the case to be heard at the CAT.

Along with Stopford’s case, in January,  the Competition and Markets Authority (CMA) began an investigation seeking to determine if Google has strategic market status in search and search advertising activities, and whether these services are delivering good outcomes for people and businesses in the UK.

The Brook case appears to be looking specifically at the cost to business arising from Google business practices that stipulate its Chrome browser and search engine are configured as the default options on Android devices and Google’s payments to Apple to ensure Google search is default on the Safari browser.

The class action also covers Google’s Search Engine Management Platform (SA360). Brook alleges that this offers better functionality and more features regarding Google’s own advertising offering than that of its competitors.

Damien Geradin, founding partner of Geradin Partners, the legal firm representing Brook, said: “This is the first claim of its kind in the UK that seeks redress for the harm caused specifically to businesses who have been forced to pay inflated prices for advertising space on Google pages.”

In the claim, Brook argues that Google has been shutting out competition in the general search and search advertising markets.

The claim argues that Google’s conduct has prevented competitors in the general search market from distributing their own search engines, which has enabled Google to maintain its dominance, leading to restricted competition in general search. Brook contests that Google has ensured that its own search platform is the only viable means of advertising to the vast majority of consumers, and ensured its dominance in search advertising.

She said: “Today, UK businesses and organisations, big or small, have almost no choice but to use Google ads to advertise their products and services. Regulators around the world have described Google as a monopoly and securing a spot on Google’s top pages is essential for visibility.

“Google has been leveraging its dominance in the general search and search advertising market to overcharge advertisers. This class action is about holding Google accountable for its unlawful practices and seeking compensation on behalf of UK advertisers who have been overcharged.”

On top of the class actions, Google is also being investigated by the CMA, which is looking at whether its Play Store requires app developers to sign up to unfair terms and conditions as a condition of distributing their apps.

Source

The iPhone 16e was just released in March, but this was apparently enough for Apple to become the world’s top smartphone vendor in the first quarter. The iPhone 16e is credited with the win for Apple, which means the affordable new iPhone 16 version is quite popular with buyers. 

I saw this coming a mile away. I told you how awesome the iPhone 16e would be all the way back when it was just a rumored device we all referred to as the iPhone SE 4. Sure, it’s more expensive than we thought, starting at $599, but it’s still a great deal. You get cutting-edge specs, tremendous battery life, an all-screen design, and Apple Intelligence. 

But you don’t have to listen to me or the quarterly smartphone sales comparisons. Google just told us how amazing the iPhone 16e is by doing something I didn’t see coming. Google posted a “Pixel 9a vs. iPhone 16e” comparison page on its website where it sells Pixel phones. 

Google is afraid. Very afraid.

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.

If you’ve been following the iPhone and the Pixel for as long as I have, you know that Google has criticized Apple’s handsets plenty of times, only to then copy Apple’s lead. It happens over and over, and Pixel phones have gotten better and better as a result.

The Pixel 9a is no different. It’s a tremendous mid-range phone, and it’s cheaper than the iPhone 16e, starting at $499 for 128GB of storage. You get Pixel 9 specs, a good camera experience, great battery life, and lots of AI thanks to Google Gemini, which is miles ahead of Apple Intelligence.

Google’s Pixel 9a vs. iPhone 16e comparison page. Image source: Google

The last thing Google should do is draw attention to the iPhone 16e by publishing a Pixel 9a comparison on its site. If you’ve got the customer here, and they’re almost ready to buy new hardware, don’t show them the competition. What if they change their mind and go for the iPhone 16e instead?

Google’s comparison page wants to answer the question: Is the Pixel 9a “the best budget phone out there?”

The answer Google is going for is “yes, the Pixel 9a is the best budget phone,” though Google isn’t exactly impartial:

Obviously, we’re big Pixel fans. So let’s step away from the subjective stuff – like how the Pixel 9a color choices are way more fun and the design is cooler – and look at the facts.

Obviously, Google then puts up a specs comparison that will mostly benefit the Android phone. That’s how these specs comparisons have gone for years. Android vendors usually win them because they’ve been beefing up the specs at a much faster pace than Apple.

Obviously, the specs detail that matters the most, the chips that power the two phones, is absent. The iPhone 16e’s A18 chip will wipe the floor with Google’s Tensor G4.

Obviously, Google also uses the “7 years of security and OS updates with Pixel Drops” category to tell you the Pixel 9a is better. First of all, iPhones don’t get Pixel Drops. Second, that G4 chip will probably have a much tougher time dealing with those Android updates the further we go. Meanwhile, Apple’s iPhone 16e should have no problem running several iOS iterations.

Obviously, Apple’s AI sucks. There’s no question about it. This is Google’s main advantage, and something the comparison is correct to point out. Gemini is simply superior to Apple Intelligence, and I say that as a longtime iPhone owner disappointed in how Apple handled its first year of AI.

Obviously, as a longtime iPhone user, I’d recommend the iPhone 16e over the Pixel 9a time and again to anyone who isn’t loyal to a smartphone brand or mobile operating system. That’s not to say the Pixel 9a is a bad phone. I’d recommend it over other mid-range phones to Android users.

What I’m getting at is that this comparison page should not exist on Google’s page. Not unless Google puts up similar comparisons with other mid-range Android phones that compete against the Pixel 9a. Samsung has a few of those, as does Nothing and virtually every Android vendor under the sun. Otherwise, it just goes to show how afraid Google is of Apple when it shouldn’t be.

Pixel phones are good, mature devices. Google can sell them without comparing them to iPhones. Obviously.

Source

Earlier this month, Anthropic announced a major new AI initiative targeting the education sector. The Claude for Education program includes an AI Learning Mode for students as well as AI tools that the entire faculty can use. Anthropic also announced partnerships with a few educational institutions.

Unsurprisingly, OpenAI didn’t wait long to respond with its own initiative, deciding to make ChatGPT Plus available to students for free for the next two months. Although OpenAI’s initiative is much more limited than Anthropic’s, it made a premium version of ChatGPT available to more students who qualify for AI-for-school deals.

It’s now Google’s turn to try to woo students with AI products, and it’s an offer students can’t and should not refuse, even if they don’t use Gemini as their primary AI companion.

Google is making Gemini Advanced available for free to students, and that includes access to the newly released Gemini 2.5 Pro model and 2TB of storage. The offer is much better than OpenAI’s as it extends through Spring 2026 as long as you subscribe by the end of June.

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.

Gemini Advanced (well, Google One) is the equivalent of ChatGPT Plus. For $20/month, you get access to Google’s latest and best models with a big twist. It’s not just AI access that you’re buying; it’s also 2TB of cloud storage that you get each month.

Sign up until June 30th, 2025, and you’ll get free access to the following tools until Spring 2026:

• Gemini 2.5 Pro – Google’s latest AI model
• Gemini Live
• Deep Research
• Canvas
• NotebookLM Plus with five times more Audio Overviews (podcasts)
• Gemini support in Google Docs, Sheets, and Slides
• Veo 2 (video generation AI)
• Whisk (image generation AI)
• 2TB of storage

All you need to do to take advantage of the offer is show your .edu email address during sign-up. That’s how you’ll verify the Gmail address that you’ll use for Gemini Advanced access.

If you already have a paid Gemini subscription (aka, you’re paying for Google One), you can switch to the free offer by canceling the current subscription and signing up for Google’s promo.

Act fast, and you’ll get up to 15 months of Gemini Advanced support for free, which would cost you up to $20/month or up to $300/month otherwise. The offer is incredible.

Again, even if you don’t use generative AI tools that often, or if you like ChatGPT more than Gemini, you still get 2TB of free storage for more than a year. That alone makes Google’s offer that much better than competing promotions.

On that note, you can and should sign up for OpenAI’s free deal, too, even if that only gives you a couple of months of free ChatGPT Plus access. Depending on where you study, you might also get Claude AI for free as part of Anthropic’s partnership with your college.

Source

Apple’s next iPhone 17 models are expected to be the last to feature a 3nm chip. This technology was first introduced with the iPhone 14 Pro, and Apple has been improving its manufacturing processes. However, it seems the company (and TSMC, of course) is almost ready to move on production to a more complex 2nm technology. Previously, BGR reported that analyst Ming-Chi Kuo predicted that some of the 2026 iPhones will move to TSMC’s 2nm chips, most likely the Pro models.

Prior to that, a Weibo leaker, Mobile Phone Chip Expert, said Apple would not only move to this new chip format but also utilize a new packaging method called Wafer-level Muti-Chip Module.

While these Weibo leakers are usually right half of the time, we were waiting for a proper source to reinforce this rumor. According to Jeff Pu’s note seen by BGR, Cupertino is indeed preparing to adopt the WMCM technology in the 2026 iPhones.

This new tech is responsible for three new enhancements: Increased flexibility, improved efficiency, and better performance. That said, WMCM packaging allows the combination of multiple chips in a package, reduces the chip’s overall size and power consumption, and even improves communication and performance.

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.

Pu agrees, “This new packaging technology minimizes overall package height and shrinks overall packaging size volume, which helps expand iPhone battery capacity. To increase AI computing performance, we expect Apple to adopt WMCM packaging, using die bonding technology to bond CPU with DRAM and Wi-Fi at a side-by-side 2.5D layout.”

In addition, Apple is expected to start producing its Wi-Fi and 5G chips internally. The first 5G modem has been available with the iPhone 16e, and the first Wi-Fi chip will likely be available with the new iPhone 17 lineups.

By 2026, Apple will already have improved versions of these chips, making the iPhone even faster and with a better battery life thanks to the seamless integration between hardware and software.

Source

A study from connectivity intelligence provider Ookla has revealed the UK’s capital lags behind the country’s large cities across key 5G performance indicators – with the gap widening – and that mobile users in London spend more time in signal not-spots with no service than residents of other UK cities, reflecting lingering coverage gaps indoors and across key transport routes.

As it made its analysis – comparing how London’s 5G performance stood up to other UK cities, including Manchester, Glasgow, Cardiff and Belfast – Ookla emphasised the global importance of London as home to one of the world’s largest and most lucrative service hubs, supporting a “vast” network of finance and technology firms.

Furthermore, it stressed how beyond its strategic time zone and English-language advantage for accessing both American and Asian markets, London’s prosperity has been founded on the availability of world-class infrastructure that facilitates doing business.

Yet the study highlighted how the city’s reputation for international competitiveness has not been matched by the quality of its telecommunications infrastructure. In particular, it has shown how London’s mobile users are experiencing frequent issues using mobile devices indoors, underground and in busy areas.

Ookla regarded London’s underperformance at the lower percentiles of measures like download speeds as particularly notable, as it said this strongly reflected the experience of mobile users in more challenging conditions such as at the network edge, during peak hours or in congested areas. Additionally, the city’s lower consistency score and weaker download and upload speeds were seen as suggesting that Londoners are more likely to encounter poor mobile performance compared with residents of other major UK cities. These problems typically manifest as poor quality of experience in everyday tasks such as web browsing, video streaming and gaming.

Worryingly for the capital city, Ookla observed that what it called London’s “marked underperformance” makes the UK unique in Western European terms – not only are the disparities between its major cities wider, but it’s also unusual for the capital to be the primary laggard.

Specifically, the study found that in the first quarter of 2025, London trailed other UK cities in 5G network consistency – a key indicator of performance at the lower end of the user experience – as well as in median download and upload speeds. Mobile users in London and Belfast experienced the weakest outcomes among UK cities, with median 5G download speeds of approximately 115 Mbps in both cities, significantly behind Glasgow’s 185 Mbps.

The study discovered that the proportion of Londoners spending the majority of their time in locations with no service (0.7%) remained higher than in other UK cities in Q1 2025, but has improved significantly from 3.7% in Q1 2023. This progress, said Ookla, reflected operator investments in network densification through small cells and the ongoing roll-out of mobile coverage across the London Underground which have together enhanced overall network availability in the capital.

Time spent on 2G networks increased, however, across several UK cities over the past year, including Birmingham and Manchester, as the advancement of the 3G sunset in the UK contributed to greater propensity for 2G fallback.

In Q1 2024, Leeds led UK cities in 5G availability, with a 21 percentage-point gap above the national average. Yet the study showed that by Q1 2025, London had taken the lead in 5G availability among major UK cities, and that gap above the national average had narrowed to 13 percentage points. This trend, said the analyst, highlighted progress in 5G network expansion in smaller UK towns and rural areas in recent months, which has moved at a faster pace than coverage improvements in larger cities.

Overall, Ookla measurements showed median 5G download speeds fell by more than 7% on average across major UK cities between Q1 2024 and Q1 2025, likely reflecting the impact of shifting network load from older technologies onto 5G, which contributed to broader improvements in overall mobile network performance in most UK cities in the same period.

Source

Businesses are under attack from all corners of the globe and while many organisations may think that nation-state threat actors would never target or be interested in them, the reality is that no-one is exempt from security threats.

Security leaders need to ensure they are staying up to speed on the latest threat intelligence, this can either be through an in-house capability or via third-party threat intel providers. Once they understand the tactics, techniques and procedures (TTPs) deployed by these threat actors, organisations can then ensure they have robust mechanisms in place to digest and act on this information to implement appropriate controls.

Organisational culture plays a key role in ensuring everyone is aware of the threats and risks posed to the business. It is vital that leaders educate users on what the most prevalent threats may look like and how to respond, this is a primary defence to protecting their business.

Social engineering remains one of the most widely used methods of attack and so implementing processes that are resistant to individual compromise is key. Using phishing resistant authentication methods, ensuring strict identity governance and control, and having a well-tested incident response capability are all crucial steps to preventing and mitigating these types of attacks.

Unfortunately, securing your own organisation is not enough and historically nation-state threat actors have taken advantage of weak third-party suppliers and supply chain governance. Having strong supply chain governance and assurance is now one of the top trends across industries and it’s critical businesses understand the dependencies and access that suppliers have.

If prevention fails, lateral movement post-compromise is one of the first actions threat actors will attempt and so endpoint detection and response, and zero-trust solutions that can prevent and detect unauthorised access are also vital.

In 2023, 1.9 billion session cookies were stolen from Fortune 1000 employees. With the session token, attackers are bypassing MFA and so it is much harder to detect and respond. Having solutions  in place as part of a zero-trust architecture to detect session token replay attempts can stop these attacks and alert to possible credential or endpoint compromise.

Ultimately, collaboration and partnership across organisations and industry will help organisations understand these threats, the risks posed by nation-state actors and more importantly allow them to work together to prevent them.

Stephen McDermid is EMEA CSO at Okta  

Source

Cyber security remained the most pressing challenge facing those in supply chain management roles during the first three months of 2025, but since the inauguration of Donald Trump in January, uncertainty over the president’s approach to tariffs has caused chaos for supply chains not just in the US, but around the world, and these two areas of risk are closely entwined.

This is according to a report from cyber and risk management consultancy West Monroe, which found that while security remains top of mind for 23% of respondents to a recent polling exercise, the impact of tariffs has surged to become the top issue for 20%, in a matter of weeks edging out factors such as geopolitical tension, material costs, the climate crisis and labour costs.

Although its fieldwork was conducted in March, prior to Trump’s so-called Liberation Day tariff announcement, West Monroe’s data shows that during Q1, a significant number of organisations in the US started making changes to their supply chains in advance.

A total of 58% said they altered their product, materials or sourcing mix, 56% altered their transportation mix, 45% altered their production schedule, 31% updated their pricing to pass increased costs to customers, and 28% altered their geographic presence. “I don’t think these are necessarily quick changes to make, but there is cyber risk if and when those changes are made,” said Christina Powers, cyber security partner at West Monroe.

Broadly, she said the need to move quickly to replace lost revenues, shifts in the supplier ecosystem and other impacts arising from the tariffs may create gaps in best practice when it comes to supply chain management.

“For example, if you’re starting to work with a different supplier – maybe they were already on your list but they weren’t a tier one supplier, you’re tapping into tier two suppliers – so maybe they went through less due diligence and less scrutiny when you were initially onboarding them,” said Powers.

“Or if you’re looking to change suppliers now, there could be a little more of a rushed diligence process being done to try to make that change more quickly,” she said. “There could be less visibility into what potential access these companies may have. From another angle, if you’re not working with a familiar contact, or not working with familiar processes, there’s a higher risk of things like impersonation attacks, whether or not that’s for financial gain or to get access to sensitive data.”

Finally, with goods potentially priced higher thanks to the tariffs, some organisations may also look to offset costs in rather more creative ways than simply passing them onto their customers. In some instances, however ill-advised this may be, this could see IT and cyber security budgets taking a hit.

“There is a risk around cyber security which is often viewed as a cost centre,” said Powers. “It is focused on value preservation and risk reduction, but it’s not necessarily value creation per se. So, there could be pushes to offset some of what organisations are having to deal with.”

But the story doesn’t end here, she said, for there are other ways in which cyber security and tariffs are coupled together.

“With a lot of the uncertainty that’s happening right now, there’s a very volatile market,” she said. “From a cyber security perspective, that could lead to incentives for individuals or groups or nation-states to look to exploit vulnerabilities or go after certain companies.

“You may see that nations that were historically friendly [to the US] have different feelings now, so there could be an increase in exploitation.

“On the data side, there could be an increase in potential espionage looking for trade secrets, intellectual property and things of that nature,” said Powers. “There are some Chinese manufacturers exploiting luxury brands and where their goods are being made, and what it takes to produce them.”

Takeaways for cyber leaders

If there’s a core message for security leaders to hold onto during this time of intense economic uncertainty and volatility, it would be not to allow the organisation to lose focus on the integrity of its supply chain arrangements.

“Now is the time to be more vigilant, not only to hold the line, but actually to increase supply chain scrutiny from a cyber perspective, because there is so much uncertainty, change, volatility and, I think, anger associated with this,” said Powers.

Source

What comes after the iPhone? Some suggest it could be AI-related. While early AI devices like Humane’s AI Pin and the R1 Rabbit were a complete failure, it seems the next batch might revolutionize the market. Besides that, if Apple’s former design chief is behind a new tech product, we should probably pay attention.

According to The Information, OpenAI is considering buying the startup founded by Jony Ive and Sam Altman, which is said to be worth at least $500 million. The io Products company develops AI-powered devices, which the publication believes could include a phone without a screen and other AI-enabled products. That said, Jony Ive could be potentially developing an AI phone, the ultimate iPhone replacement.

Still, while people familiar with the matter insist it is “not a phone,”—of course it isn’t—the idea behind it could be as revolutionary as the first iPhone. Still, the report highlights Jony Ive and Altman’s collaboration is still in the early stages, without a product concept in hand.

What makes this project even more enticing is that Laurene Powell Jobs’ Emerson Collective also funds the venture. The startup even includes former Apple designers like Tang Tan and Evans Hankey, who worked with Ive on the iPhone.

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.

At the end of the day, the full acquisition could become a partnership, where OpenAI would get the technology and engineering team while retaining its AI capabilities. At the same time, Jony Ive’s LoveFrom studio would take care of the design part.

Since he left Apple, Jony Ive hasn’t worked on new hardware. However, its design kept making the headlines with the King of England, Ferrari’s partnership, Moncler, and so on.

Reports about Jony Ive working with OpenAI aren’t new. At least since 2023, BGR has reported that Apple’s former design chief has been working on the future iPhone AI. However, with OpenAI planning to buy the startup, the company may be about to arrive at a breakthrough tech point, which means we could soon see hardware from OpenAI that is as innovative as the first iPhone.

Source

CIOs have been under immense pressure for some time to deliver successful digital initiatives while navigating budget constraints and increasing demands from senior executives. A recent Gartner survey reveals that 92% of CIOs anticipate integrating artificial intelligence (AI) into their organisations by 2025, yet 49% struggle to assess and showcase the technology’s value. Are we going round in circles here?

Amid these challenges, small language models (SLMs) have emerged as a compelling solution, promising lower-cost and more secure AI capabilities that can fit with strategic priorities. So much about SLMs makes sense.

“The AI community has been actively exploring small language models like Mistral Small and DeepSeek R1,” says Amer Sheikh, chief data scientist at BearingPoint. “These models have seen significant traction, as evidenced by the number of downloads on Hugging Face. Their popularity stems from their ability to trade off accuracy, speed and cost-effectiveness.”

Adding intelligence at the edge

And that’s the key point. It is a trade-off – but one that is clearly worth making. SLMs, by their very nature, offer a practical alternative for organisations seeking to implement AI without the overheads associated with large language models (LLMs). They are also driving the next wave of edge AI adoption, enabling AI models to run on smartphones, internet of things (IoT) devices and industrial systems without relying on cloud infrastructure.

“Small models open up the possibility to push execution to the edge,” says Peter van der Putten, director of the AI Lab at Pegasystems and assistant professor of AI at Leiden University. “This could mean running on high-end smartphones, IoT devices such as cameras and, with proper consent, unlocking completely new data sources to learn from that are currently not available on the open internet.”

Despite the promise, real-world applications of SLMs in mobile and IoT devices remain in the early stages. Some practical implementations include DeepSeek’s R1 model, which has been integrated into Chinese automakers’ infotainment systems (such as Geely), and Phi-3, a small model designed for mobile AI applications. In education, Stanford’s Smile Plug uses small AI models to deliver interactive learning experiences on Raspberry Pi devices without internet connectivity. These examples demonstrate the growing potential of SLMs.

“SLMs can and are being deployed in a number of industries where there is a requirement for specific domain knowledge,” adds Sheikh, highlighting their use in customer service chatbots, virtual assistants and text summarisation.

Unlike LLMs, which require vast computational power and cloud resources, SLMs can run locally, cutting costs and mitigating security risks, hence their suitability for enhancing edge device intelligence. “There is a massive reduction in inference costs. However, there will be small costs for fine-tuning and self-hosting,” he adds.

SLMs can be augmented with smaller, more focused datasets, says Isabel Al-Dhahir, principal analyst at GlobalData. “Employing SLMs circumvents several challenges associated with general-purpose LLMs, including computational power requirements, exorbitant costs and insufficient domain knowledge.”

This ability to focus on precise, industry-specific use cases is why regulated sectors such as telecoms, accounting and law are adopting SLMs more readily.

“We have seen SLMs for professional services in dealing with accounting regulation, telecoms regulation, and various on-device applications and home automation,” Al-Dhahir adds.

With retrieval augmented generation (RAG) techniques, businesses can further refine and enhance the accuracy of these models within their specific domains.

Security key focus for industry growing LLM-weary

Beyond cost, security remains a major factor, especially within edge devices. According to Saman Nasrolahi, principal at InMotion Ventures (Jaguar Land Rover’s investment arm), this is where SLMs are also ticking a few boxes.

Much of the fear around LLMs is associated with a lack of transparency as to what is going on behind the scenes in terms of data collation and analytics. SLMs are the on-premise version of the generative artificial intelligence (GenAI) world.

“In addition to cost reduction, this approach also makes them far more secure and less vulnerable to data breaches as data does not need to leave an organisation’s borders,” says Nasrolahi.

This capability is particularly crucial for the healthcare, financial services and legal sectors, where regulatory compliance and data protection are paramount.

“Approximately one-third of all cyber security attacks occur when data is shared with an external vendor. By keeping data on-site, SLMs can reduce the attack surface and enterprise vulnerabilities,” Nasrolahi adds.

In a time when businesses are increasingly concerned about data sovereignty and compliance, the ability to localise AI processing is surely a significant advantage.

Andrew Bolster, senior research and development manager (data science) at Black Duck, adds that the portability of SLMs, at least compared with “the juggernauts of GPT-4, Claude, or even Llama”, makes them well suited to edge deployment. Security, cost and functionality are attractive propositions.

“SLMs operating on edge devices mean users’ data doesn’t have to leave the device to contribute to an intelligent response or action while potentially improving latency and performance, making intelligent operations feel more ‘relevant’ and ‘snappy’ while protecting users’ privacy,” he says.

With advances in custom chipsets to support these kinds of workloads, the power, memory and performance requirements of SLMs can now be found in most laptops and mid-tier mobile phones, allowing service platforms to shift more intelligence closer to the end user. This ability to process data locally on laptops, mobile devices and industrial IoT systems makes SLMs particularly valuable for low-latency applications, security-sensitive industries and environments with limited internet access. 

Jeff Watkins, chief technology officer (CTO) at CreateFuture, adds that SLMs “can run locally on laptops, desktop computers, smartphones, or even IoT devices. They range in sizes and capabilities – from ones that can run on compact devices to ones that begin to challenge the latest MacBook Pro models”.

With lower costs, enhanced security and the ability to function efficiently on existing hardware, SLMs present an increasingly strategic option for businesses. But as with any emerging technology, challenges remain. Hallucinations, biases and the need for fine-tuning mean it requires careful implementation.

“Hallucinations are still a problem for SLMs, similar to LLMs. Though, more specialised models tend to be less susceptible to these issues,” says Nasrolahi.

Lower the energy, lower the cost, the more mobile it becomes

Another key driver for the adoption of SLMs in edge devices is their ability to operate with lower energy consumption while also reducing cloud dependency. “SLMs are less energy-intensive, making them cheaper, better for the environment, and often small enough to run locally on edge compute such as your mobile or PC without the need for an internet connection,” says Silvia Lehnis, consulting director for data and AI at UBDS Digital.

The environmental and operational cost benefits make SLMs particularly appealing for businesses aiming to reduce their AI carbon footprint while maintaining data security. “Running the model locally without internet access can also have data privacy advantages, as your data is not being shared with an online application for central logging and monitoring, making it suitable for more sensitive use cases,” adds Lehnis.

It’s a recurring theme. This growing awareness that SLMs can enable a shift away from one-size-fits-all LLMs toward more focused, cost-efficient AI models should change how enterprises think about GenAI use. It could have a broader impact on IT buying, certainly in terms of how CIOs think strategically about what is and isn’t possible with GenAI.

Deloitte’s Tech Trends 2025 report suggests enterprises are now considering SLMs and open source options for the ability to train models on smaller, more accurate datasets. It’s a recognition that size isn’t everything, but accuracy and relevance is, aligning any AI deployments with operational objectives.

The trajectory of AI adoption indicates a growing preference for models that balance performance with operational practicality, but there is also a growing desire for more edge computing, real-time and strategically relevant functionality.

Interestingly, back in 2017, Gartner predicted this would happen, claiming that by this year, 75% of enterprise-generated data would be created and processed outside traditional centralised datacentres or the cloud. And that was before we knew anything about SLMs and their role.

So, what does this mean for the future of SLMs and edge computing devices? Certainly, they will have a significant role to play as enterprises see AI on their terms but also to enable differentiation. That will become the new challenge for CIOs – how to get the best out of GenAI to make a big impact on business performance. Angles for this can come from a number of directions – it really depends on the organisation and the industry.

The rise of SLMs is not just about cost savings or security – it’s about AI differentiation. As Jarrod Vawdrey, field chief data scientist at Domino Data Lab, points out, SLMs are already reshaping healthcare, finance and defence, allowing on-device AI to reduce latency, protect sensitive data and enhance real-time decision-making.

“SLMs deployed on medical devices enable real-time patient monitoring and diagnostic assistance,” he notes, while financial institutions are leveraging SLMs for fraud detection and anti-money laundering compliance.

For CIOs, the challenge is shifting. How do you harness GenAI to make a significant impact on business performance? The answer lies in adapting AI models to industry-specific needs – something SLMs are uniquely positioned to do. The next few years will see enterprises move beyond generic AI models, focusing instead on hyper-relevant, domain-trained AI that drives differentiation and competitive advantage. If anything is going to push edge computing into the mainstream, it’s small language models.

Source