Gamers who are constantly on the lookout for a new screen will notice when a model comes up that provides excellent visuals at a reasonable price. The INNOCN 27″ QD-OLED 2K (model GA27S1Q) is a prime example, priced at $399.98 (was $450). Once out of the box, the stand snaps into place without the need for tools, and you have full movement in all directions, including height, tilt, swivel, and pivot. So, if you’re the type of person who enjoys switching between working at a desk and gaming on the sofa, you can find the perfect angle.

The images on-screen are noticeably vibrant right away, with black areas remaining deep / dark rather than washing out to gray, bringing the highlights and colors to life in each scene. The panel also covers almost all of the colors required for modern games and media, so reds and greens appear nice and vibrant with no dull areas, and the animation remains very clean even when things get really fast. With a 280Hz refresh rate that can reach 280 frames per second and a response time measured in thousandths of a second, fast-moving objects maintain sharp edges and prevent blurring that occurs on slower panels. If you’re a die-hard gamer, you’ll notice the difference in quick turns and abrupt adversary movements, whereas casual sessions simply feel more responsive overall.

Sale

INNOCN 27″ QD-OLED 2K QHD 2560 x 1440P 280Hz 240Hz PC Computer Gaming Console Monitor, G-Sync Compatible…

• Experience Ultimate Gaming Visual Clarity: This 27-inch QD-OLED gaming monitor delivers stunning 1440p resolution with perfect blacks and vibrant…
• Dominate with Blur-Free 280Hz Speed: Gain the competitive edge with a blistering 280Hz refresh rate and near-instantaneous 0.03ms response time. Enjoy…
• Next-Gen QD-OLED Visual Fidelity: Witness breathtaking contrast and rich colors powered by QD-OLED technology. Enjoy immersive PC gaming and HDR…

Connections are rather comprehensive for a monitor at this price point, with two HDMI 2.1 connections capable of handling consoles and newer graphics cards at full speed, as well as a pair of DisplayPort 1.4 inputs for further versatility if you have a desktop system. The built-in speakers will suffice for brief checks and the odd thing, but most people prefer to plug in headphones for better sound during extended playback.

This monitor has features for both comfort and lifespan. It has low blue light and flicker-free settings to lessen eye strain if you stare at it for an extended period of time, as well as some useful routines that look for static images and adjust brightness to prevent permanent markings from appearing. Don’t worry about the power drain; it’s rather low at roughly 65 watts, so the monitor runs cool and won’t put too much burden on your outlet.

One other advantage is that its slim bezels keep your focus on the image, and the rear has some modest illumination that adds a stylish touch to your setup without drawing too much attention. Overall, for anyone looking at monitors of this size and resolution, this one demonstrates that you don’t have to trade quality for a reasonable price. Give it a few hours, and you’ll see why the word is spreading so quickly, as the mix of crystal-clear visuals and seamless pace makes it a true winner.

The race to transition online security protocols to ones that can’t be cracked by a quantum computer is already on. The algorithms that are commonly used today to protect data online—RSA and elliptic curve cryptography—are uncrackable by supercomputers, but a large enough quantum computer would make quick work of them. There are algorithms secure enough to be out of reach for both classical and future quantum machines, called post-quantum cryptography, but transitioning to these is a work in progress.

Late last month, the team at Google Quantum AI published a whitepaper that added significant urgency to this race. In it, the team showed that the size of a quantum computer that would pose a cryptographic threat is approximately twenty times smaller than previously thought. This is still far from accessible to the quantum computers that exist today: the largest machines currently consist of approximately 1,000 quantum bits, or qubits, and the whitepaper estimated that about 500 times as much is needed. Nonetheless, this shortens the timeline to switch over to post-quantum algorithms.

The news had a surprising beneficiary: obscure cryptocurrency Algorand jumped 44% in price in response. The whitepaper called out Algorand specifically for implementing post-quantum cryptography on their blockchain. We caught up with Algorand’s chief scientific officer and professor of computer science and engineering at the University of Michigan, Chris Peikert, to understand how this announcement is impacting cryptography, why cryptocurrencies are feeling the effects, and what the future might hold. Peikert’s early work on a particular type of algorithm known as lattice cryptography underlies most post-quantum security today.

IEEE Spectrum: What is the significance of this Google Quantum AI whitepaper?

Peikert: The upshot of this paper is that it shows that a quantum computer would be able to break some of the cryptography that is most widely used, especially in blockchains and cryptocurrencies, with much, much fewer resources than had previously been established. Those resources include the time that it would take to do so and the number of qubits (or quantum bits) that it would have to use.

This cryptography is very central to not just cryptocurrencies but more broadly, to cryptography on the internet. It is also used for secure web connections between web browsers and web servers. Versions of elliptic curve cryptography are used in national security systems and military encryption. It’s very prevalent and pervasive in all modern networks and protocols.

And not only was this paper improving the algorithms, but there was also a concurrent paper showing that the hardware itself was substantially improved. The claim here was that the number of physical qubits needed to achieve a certain kind of logical qubit was also greatly reduced. These two kinds of improvements are compounding upon each other. It’s a kind of a win-win situation from the quantum computing perspective, but a lose-lose situation for cryptography.

IEEE Spectrum: What do Google AI’s findings mean for cryptocurrencies and the broader cybersecurity ecosystem?

Peikert: There’s always been this looming threat in the distance of quantum computers breaking a large fraction of the cryptography that’s used throughout the cryptocurrency ecosystem. And I think what this paper did was really the loudest alarm yet that these kinds of quantum attacks might not be as far off as some have suspected, or hoped, in recent years. It’s caused a re-evaluation across the industry, and a moving up of the timeline for when quantum computers might be capable of breaking this cryptography.

When we think about the timelines and when it’s important to have completed these transitions [to post-quantum cryptography], we also need to factor in the unknown improvements that we should expect to see in the coming years. The science of quantum computing will not stay static, and there will be these further breakthroughs. We can’t say exactly what they will be or when they will come, but you can bet that they will be coming.

IEEE Spectrum: What is your guess on if or when quantum computers will be able to break cryptography in the real world?

Peikert: Instead of thinking about a specific date when we expect them to come, we have to think about the probabilities and the risks as time goes on. There have been huge breakthrough developments, including not only this paper, but also some last year. But even with these, I think that the chance of a cryptographic attack by quantum computers being successful in the next three years is extremely low, maybe less than a percent. But then, as you get out to several years, like 5, 6, or 10 years, one has to seriously consider a probability, maybe 5% or 10% or more. So it’s still rather small, but significant enough that we have to worry about the risk, because the value that is protected by this kind of cryptography is really enormous.

The US government has put 2035 as its target for migrating all of the national security systems to post quantum cryptography. That seems like a prudent date, given the timelines that it takes to upgrade cryptography. It’s a slow process. It has to be done very deliberately and carefully to make sure that you’re not introducing new vulnerabilities, that you’re not making mistakes, that everything still works properly. So, you know, given the outlook for quantum computers on the horizon, it’s really important that we prepare now, or ideally, yesterday, or a few years ago, for that kind of transition.

IEEE Spectrum: Are there significant roadblocks you see to industrial adoption of post-quantum cryptography going forward?

Peikert: Cryptography is very hard to change. We’ve only had one or maybe two major transitions in cryptography since the early 1980s or late 1970s when the field first was invented. We don’t really have a systematic way of transitioning cryptography.

An additional challenge is that the performance tradeoffs are very different in post-quantum cryptography than they are in the legacy systems. Keys and cipher texts and digital signatures are all significantly larger in post-quantum cryptography, but the computations are actually faster, typically. People have optimized cryptography for speed in the past, and we have very good fast speeds now for post-quantum cryptography, but the sizes of the keys are a challenge.

Especially in blockchain applications, like cryptocurrencies, space on the blockchain is at a premium. So it calls for a reevaluation in many applications of how we integrate the cryptography into the system, and that work is ongoing. And, the blockchain ecosystem uses a lot of advanced cryptography, exotic things like zero-knowledge proofs. In many cases, we have rudimentary constructions of these fancy cryptography tools from post-quantum type mathematics, but they’re not nearly as mature and industry ready as the legacy systems that have been deployed. It continues to be an important technical challenge to develop post-quantum versions of these very fancy cryptographic schemes that are used in cutting edge applications.

IEEE Spectrum: As an academic cryptography researcher, what attracted you to work with a cryptocurrency, and Algorand in particular?

Peikert: My former PhD advisor is Silvio Micali, the inventor of Algorand. The system is very elegant. It is a very high performing blockchain system and it uses very little energy, has fast transaction finalization, and a number of other great features. And Silvio appreciated that this quantum threat was real and was coming, and the team approached me about helping to improve the Algorand protocol at the basic levels to become more post-quantum secure in 2021. That was a very exciting opportunity, because it was a difficult engineering and scientific challenge to integrate post-quantum cryptography into all the different technical and cryptographic mechanisms that were underlying the protocol.

IEEE Spectrum: What is the current status of post-quantum cryptography in Algorand, and blockchains in general?

Peikert: We’ve identified some of the most pressing issues and worked our way through some of them, but it’s a many-faceted problem overall. We started with the integrity of the chain itself, which is the transaction history that everybody has to agree upon.

Our first major project was developing a system that would add post-quantum security to the history of the chain. We developed a system called state proofs for that, which is a mixture of ordinary post-quantum cryptography and also some more fancy cryptography: It’s a way of taking a large number of signatures and digesting them down into a much smaller number of signatures, while still being confident that these large number of signatures actually exist and are properly formed. We also followed it with other papers and projects that are about adding post-quantum cryptography and security to other aspects of the blockchain in the Algorand ecosystem.

It’s not a complete project yet. We don’t claim to be fully post-quantum secure. That’s a very challenging target to hit, and there are aspects that we will continue to work on into the near future.

IEEE Spectrum: In your view, will we adopt post-quantum cryptography before the risks actually catch up with us?

Peikert: I tend to be an optimist about these things. I think that it’s a very good thing that more people in decision making roles are recognizing that this is an important topic, and that these kinds of migrations have to be done. I think that we can’t be complacent about it, and we can’t kick the can down the road much longer. But I do see that the focus is being put on this important problem, so I’m optimistic that most important systems will eventually have good either mitigations or full migrations in place.

But it’s also a point on the horizon that we don’t know exactly when it will come. So, there is the possibility that there is a huge breakthrough, and we have many fewer years than we might have hoped for, and that we don’t get all the systems upgraded that we would like to have fixed by the time quantum computers arrive.

Microsoft assigned CVE-2026-21520, a CVSS 7.5 indirect prompt injection vulnerability, to Copilot Studio. Capsule Security discovered the flaw, coordinated disclosure with Microsoft, and the patch was deployed on January 15. Public disclosure went live on Wednesday.

That CVE matters less for what it fixes and more for what it signals. Capsule’s research calls Microsoft’s decision to assign a CVE to a prompt injection vulnerability in an agentic platform “highly unusual.” Microsoft previously assigned CVE-2025-32711 (CVSS 9.3) to EchoLeak, a prompt injection in M365 Copilot patched in June 2025, but that targeted a productivity assistant, not an agent-building platform. If the precedent extends to agentic systems broadly, every enterprise running agents inherits a new vulnerability class to track. Except that this class cannot be fully eliminated by patches alone.

Capsule also discovered what they call PipeLeak, a parallel indirect prompt injection vulnerability in Salesforce Agentforce. Microsoft patched and assigned a CVE. Salesforce has not assigned a CVE or issued a public advisory for PipeLeak as of publication, according to Capsule’s research.

What ShareLeak actually does

The vulnerability that the researchers named ShareLeak exploits the gap between a SharePoint form submission and the Copilot Studio agent’s context window. An attacker fills a public-facing comment field with a crafted payload that injects a fake system role message. In Capsule’s testing, Copilot Studio concatenated the malicious input directly with the agent’s system instructions with no input sanitization between the form and the model.

The injected payload overrode the agent’s original instructions in Capsule’s proof-of-concept, directing it to query connected SharePoint Lists for customer data and send that data via Outlook to an attacker-controlled email address. NVD classifies the attack as low complexity and requires no privileges.

Microsoft’s own safety mechanisms flagged the request as suspicious during Capsule’s testing. The data was exfiltrated anyway. The DLP never fired because the email was routed through a legitimate Outlook action that the system treated as an authorized operation.

Carter Rees, VP of Artificial Intelligence at Reputation, described the architectural failure in an exclusive VentureBeat interview. The LLM cannot inherently distinguish between trusted instructions and untrusted retrieved data, Rees said. It becomes a confused deputy acting on behalf of the attacker. OWASP classifies this pattern as ASI01: Agent Goal Hijack.

The research team behind both discoveries, Capsule Security, found the Copilot Studio vulnerability on November 24, 2025. Microsoft confirmed it on December 5 and patched it on January 15, 2026. Every security director running Copilot Studio agents triggered by SharePoint forms should audit that window for indicators of compromise.

PipeLeak and the Salesforce split

PipeLeak hits the same vulnerability class through a different front door. In Capsule’s testing, a public lead form payload hijacked an Agentforce agent with no authentication required. Capsule found no volume cap on the exfiltrated CRM data, and the employee who triggered the agent received no indication that data had left the building. Salesforce has not assigned a CVE or issued a public advisory specific to PipeLeak as of publication.

Capsule is not the first research team to hit Agentforce with indirect prompt injection. Noma Labs disclosed ForcedLeak (CVSS 9.4) in September 2025, and Salesforce patched that vector by enforcing Trusted URL allowlists. According to Capsule’s research, PipeLeak survives that patch through a different channel: email via the agent’s authorized tool actions.

Naor Paz, CEO of Capsule Security, told VentureBeat the testing hit no exfiltration limit. “We did not get to any limitation,” Paz said. “The agent would just continue to leak all the CRM.”

Salesforce recommended human-in-the-loop as a mitigation. Paz pushed back. “If the human should approve every single operation, it’s not really an agent,” he told VentureBeat. “It’s just a human clicking through the agent’s actions.”

Microsoft patched ShareLeak and assigned a CVE. According to Capsule’s research, Salesforce patched ForcedLeak’s URL path but not the email channel.

Kayne McGladrey, IEEE Senior Member, put it differently in a separate VentureBeat interview. Organizations are cloning human user accounts to agentic systems, McGladrey said, except agents use far more permissions than humans would because of the speed, the scale, and the intent.

The lethal trifecta and why posture management fails

Paz named the structural condition that makes any agent exploitable: access to private data, exposure to untrusted content, and the ability to communicate externally. ShareLeak hits all three. PipeLeak hits all three. Most production agents hit all three because that combination is what makes agents useful.

Rees validated the diagnosis independently. Defense-in-depth predicated on deterministic rules is fundamentally insufficient for agentic systems, Rees told VentureBeat.

Elia Zaitsev, CrowdStrike’s CTO, called the patching mindset itself the vulnerability in a separate VentureBeat exclusive. “People are forgetting about runtime security,” he said. “Let’s patch all the vulnerabilities. Impossible. Somehow always seem to miss something.” Observing actual kinetic actions is a structured, solvable problem, Zaitsev told VentureBeat. Intent is not. CrowdStrike’s Falcon sensor walks the process tree and tracks what agents did, not what they appeared to intend.

Multi-turn crescendo and the coding agent blind spot

Single-shot prompt injections are the entry-level threat. Capsule’s research documented multi-turn crescendo attacks where adversaries distribute payloads across multiple benign-looking turns. Each turn passes inspection. The attack becomes visible only when analyzed as a sequence.

Rees explained why current monitoring misses this. A stateless WAF views each turn in a vacuum and detects no threat, Rees told VentureBeat. It sees requests, not a semantic trajectory.

Capsule also found undisclosed vulnerabilities in coding agent platforms it declined to name, including memory poisoning that persists across sessions and malicious code execution through MCP servers. In one case, a file-level guardrail designed to restrict which files the agent could access was reasoned around by the agent itself, which found an alternate path to the same data. Rees identified the human vector: employees paste proprietary code into public LLMs and view security as friction.

McGladrey cut to the governance failure. “If crime was a technology problem, we would have solved crime a fairly long time ago,” he told VentureBeat. “Cybersecurity risk as a standalone category is a complete fiction.”

The runtime enforcement model

Capsule hooks into vendor-provided agentic execution paths — including Copilot Studio’s security hooks and Claude Code’s pre-tool-use checkpoints — with no proxies, gateways, or SDKs. The company exited stealth on Wednesday, timing its $7 million seed round, led by Lama Partners alongside Forgepoint Capital International, to its coordinated disclosure.

Chris Krebs, the first Director of CISA and a Capsule advisor, put the gap in operational terms. “Legacy tools weren’t built to monitor what happens between prompt and action,” Krebs said. “That’s the runtime gap.”

Capsule’s architecture deploys fine-tuned small language models that evaluate every tool call before execution, an approach Gartner’s market guide calls a “guardian agent.”

Not everyone agrees that intent analysis is the right layer. Zaitsev told VentureBeat during an exclusive interview that intent-based detection is non-deterministic. “Intent analysis will sometimes work. Intent analysis cannot always work,” he said. CrowdStrike bets on observing what the agent actually did rather than what it appeared to intend. Microsoft’s own Copilot Studio documentation provides external security-provider webhooks that can approve or block tool execution, offering a vendor-native control plane alongside third-party options. No single layer closes the gap. Runtime intent analysis, kinetic action monitoring, and foundational controls (least privilege, input sanitization, outbound restrictions, targeted human-in-the-loop) all belong in the stack. SOC teams should map telemetry now: Copilot Studio activity logs plus webhook decisions, CRM audit logs for Agentforce, and EDR process-tree data for coding agents.

Paz described the broader shift. “Intent is the new perimeter,” he told VentureBeat. “The agent in runtime can decide to go rogue on you.”

VentureBeat Prescriptive Matrix

The following matrix maps five vulnerability classes against the controls that miss them, and the specific actions security directors should take this week.

What this means for 2026 security planning

Microsoft’s CVE assignment will either accelerate or fragment how the industry handles agent vulnerabilities. If vendors call them configuration issues, CISOs carry the risk alone.

Treat prompt injection as a class-level SaaS risk rather than individual CVEs. Classify every agent deployment against the lethal trifecta. Require runtime enforcement for anything moving to production. Brief the board on agent risk the way McGladrey framed it: as business risk, because cybersecurity risk as a standalone category stopped being useful the moment agents started operating at machine speed.

Google made an unexpected cameo on Macs with the launch of a native Gemini app. What’s even more interesting (and a bit funny) is that the app arrived at Apple’s long-promised Siri upgrade (and a rumored standalone app for the voice assistant). 

The free app is available on macOS 15 and above. Though the app isn’t available on the App Store (yet), you can download it from Google’s official landing page.

What can the Gemini Mac app actually do?

Quite a bit, actually. Once you install the app, you can summon Gemini by pressing Option + Space keys. Doesn’t matter where you are and what you’re doing; using the shortcut opens a quick-access mini chat overlay. Don’t press the wrong key (Command), or you’ll end up invoking the Spotlight search bar. 

You can open the full Gemini interface by pressing Option + Shift + Space. Further, the app includes built-in tools for generating images and videos, analyzing content on your screen (including documents, spreadsheets, and images), and understanding files. Of course, you can talk to the Gemini AI assistant.

The list of available tools includes Canvas, Deep Research, NotebookLM integration, and Personal Intelligence, which taps into your connected Google apps, including Gmail, Photos, Calendar, etc., to fetch relevant information for you. 

Why does this matter for everyday Mac users?

If you don’t know this already, Gemini is among the last AI services to have launched a dedicated Mac app. Other giants — OpenAI, Anthropic, and Perplexity — have had Mac apps for quite some time. 

For Mac users who’ve been using Gemini in Chrome or Safari, the native app is a welcome upgrade. The powerful, context-aware AI is now one keyboard shortcut away on your Mac. 

By establishing Gemini on macOS now, Google secures mindshare and daily habit formation before Apple can actually flip the switch with the dedicated Siri app later this year. 

• Fake iCloud deletion emails are pressuring Apple users into dangerous clicks
• Poor grammar in iCloud alerts remains a clear sign of fraud
• Clicking fake iCloud upgrade links can expose banking and personal data

A wave of deceptive emails is attempting to pressure Apple users into believing their iCloud data is at immediate risk of deletion, using increasingly aggressive language to force quick reactions.

The messages often claim a user’s storage limits have been exceeded or that an account has been blocked, followed by threats that photos and videos will be permanently erased on a specified date.

While the Trump administration’s extremely aggressive, thoroughly bigoted attempts to eliminate as many non-white people from this country as possible have resulted in some periodic push back from law enforcement officials, we can never forget that federal law enforcement officers are still just law enforcement officers. And, more often than not, they’ll always have the support of their brothers in blue, even though most federal officers prefer camo and face masks these days.

Law enforcement is self-selecting. The people who feel drawn to law enforcement are generally the last people you would want to become law enforcement officers. It’s rarely about being given the chance to serve, protect, and be an active part of your community. It’s almost always about having a badge, a gun, and accountability that’s inversely proportional to the amount of power you immediately obtain.

So, it comes as no surprise that cops who shouldn’t have any skin in the anti-ICE game are stepping up to punish people for daring to criticize the actions of those federal officers. And there’s probably a bit of backlash involved here as well, as this following report details the actions of California law enforcement officers who (one assumes) aren’t thrilled the state’s residents have managed to reclaim much of the power that has always been owed to the people.

Despite the administration’s on/off surges in “blue” states, the furor over ICE and its actions hasn’t died down, not even in California, where the administration rolled out its martial law beta test. At first, it was easy to pretend people protesting ICE were “woke radicals” or “antifa” or “paid organizers” or “lazy trans everywhere college students” or whatever. But it just kept going and expanding, clearly demonstrating a significant portion of the population wasn’t on board with roving kidnapping squads and murders of activists by jumpy recruits recently introduced to the wholly domestic War on Migrants.

Now that it’s everyone rather than just the usual left-wing agitprop cliches federal and local officers expected to confront during protests, cops in California are deciding it’s time to start arresting everyone.

The Clovis Police Department on Tuesday referred Alfred Aldrete, 41, for one count of contributing to the delinquency of a minor for his role in a February high school student walkout. 

“During the investigation, Aldrete was identified as being present during the walkout and allegedly involved in directing student activity and entering the roadway, which impacted traffic flow,” Clovis police said in a press release. “Investigators also identified Aldrete as being present during a separate student gathering in Clovis on Feb. 5 that occurred outside of school hours.”

Yep, that’s what the Clovis PD actually did: it equated an adult ensuring students made it to their planned protest safely with the sort of horrors — harboring runaways, providing drugs and alcohol to minors, etc. — people usually associate with the crime of “contributing to the delinquency of a minor.” Those would be the sorts of crimes actually prosecuted by county prosecutors under this statute.

This stat may explain why the Clovis PD thought it should explore the fringes of this statute for the sole purpose of punishing someone for speech they (and they people they serve, apparently) don’t care for:

[C]lovis, population 128,000, where Donald Trump won every precinct in the 2024 presidential election — some with more than 70% of the vote. 

That tracks. Fortunately, it doesn’t track as far as the District Attorney’s office:

A representative for Fresno County District Attorney Lisa Smittcamp in a written statement said prosecutors would not file charges against Aldrete.

Hooray for prosecutorial discretion, but in the non-pejorative sense! It’s an unexpected twist that only makes this further twist even more inexplicable:

Within a day of the walkout, Clovis police said they were considering charges against up to six adults under Section 272 of the California Penal Code, which is most often used to prevent chronic truancy. The Los Angeles Police Department has also said it’s considering charges against people who joined immigration-related protests under the same penal code section. 

At the beginning of Trump’s first martial law-esque surge, the LAPD (and the Los Angeles Sheriffs Department) were opposed to the insertion of National Guard units and other federal officers into the mix. Stating that they were capable of handling whatever minimal “violent protests” they had actually encountered, law enforcement officials made it clear that this federal interloping would only make a manageable problem unmanageable.

More than a year later, the LAPD has flipped the script from blue to red, declaring it’s willing to charge students for truancy (along with the adults who assist them) for participating in walkout that, at best, lasts a few hours. It’s not like these kids are quitting school to pursue a career in protesting. And it’s not like these adults are harming kids by helping them engage fully with their First Amendment rights.

It’s one thing to be the main characters in a pro-Trump town. It’s quite another to be part of the second-largest police force in the United States and decide it’s worth your time, money, and attention to punish people for peacefully protesting. Fuck right off, LAPD. And take the Clovis PD with you.

Filed Under: 1st amendment, acab, alfred aldrete, california, clovis pd, free speech, ice, lapd, mass deportation

AI advancements allow workers to reduce repetitive work and ‘increase velocity’, Spiegel said.

Snap is laying off 16pc of its workforce to cut costs and veer towards long-term profitability. The Snapchat parent company is cutting around 1,000 employees, including 300 open roles.

In a memo sent to employees today (15 April), company CEO Evan Spiegel said that Snap is prioritising investments with the potential for long-term growth. He said that AI advancements allow workers to reduce repetitive work and “increase velocity”.

The layoffs are expected to reduce the company’s annual costs by more than $500m by the second half of the year, according to Spiegel. Snap shares rose more than 7.75pc in pre-market trading, but have overall been down nearly 30pc since last year.

Snapchat, alongside other major social media platforms, has been under regulatory scrutiny over the past few years over issues surrounding child safety and access to content. The platform has been banned for those under 16 in Australia.

Snap last laid off 500 jobs in 2024. At the time, the company said that the layoffs would “reduce hierarchy and promote in-person collaboration”. Two years prior, it cut around 20pc of the company to improve business performance.

Spiegel is the latest in a growing list of company leaders linking layoffs to AI advancements. In his memo, he said small teams leveraging AI tools have already had a positive impact on Snap’s ad platform performance.

In February, Jack Dorsey cut 4,000 jobs at Block in preference for AI tools and flatter teams. Since then, Atlassian cut 10pc of its workforce, Meta laid off several hundred, and Oracle cut thousands, reportedly over AI.

Dorsey, at the time, said that a “majority of companies” will reach similar conclusions around smaller teams, and make similar structural changes “within the next year”.

Journalist Alex Heath, meanwhile, has reported that Snap’s $400m deal with Perplexity has also been axed.

Announced last November, the deal would have seen Perplexity deploy its conversational search tool into Snapchat. The one-year partnership was expected to rebrand Snapchat into a platform where AI companies could connect with the platform’s community.

Don’t miss out on the knowledge you need to succeed. Sign up for the Daily Brief, Silicon Republic’s digest of need-to-know sci-tech news.

The Asus VivoBook and ZenBook laptops are quite the hotcakes in the Indian market, simply because they strike the right balance between portability and performance. Keeping up that momentum, the Taiwanese laptop maker has opened pre-orders for its latest premium Zenbook lineup in India, introducing a range of new laptops focused on design, portability, and AI-powered performance. The lineup includes the Zenbook S14, Zenbook DUO, Zenbook A14, and the upcoming Zenbook A16, alongside refreshed Vivobook models. Prices for the ZenBook series start at ₹1,79,990, while the Vivobook lineup begins at ₹98,990. Here’s everything you need to know about them.

What’s New with the ZenBooks & VivoBooks?

ASUS is doubling down on its “Design You Can Feel” philosophy with this launch. One of the key highlights is Ceraluminum, a proprietary material that aims to combine durability with a lightweight, premium finish. The flagship Zenbook S14, for instance, features an ultra-slim profile of around 1.1cm and weighs roughly 1.2kg, making it highly portable. It also gets a 14-inch 3K OLED display and a claimed battery life of up to 27 hours. Under the hood, the ZenBook S14 series packs Ultra Series 3 processors, with the highest tier going to the Ultra 9.

Meanwhile, the newest version of my favorite ZenBook Duo takes things to another level by packing dual 14-inch 3K 144Hz ASUS Lumina Pro OLED touch displays. It’s powered by Intel’s latest Core Ultra 7-series processors and offers 32 hours of claimed battery life. Lastly, there’s the Zenbook A Series (A14 & A16), which targets more casual, yet premium users. It’s made from the same Ceraluminum material and focuses primarily on portability, weighing under 1 kg. On the power side, the ZenBook A series uses the Snapdragon X2 series processors. While this should pay pretty big dividends in the battery life department, we have yet to test the laptops, so stay tuned for a full review.

The next-gen VivoBook classic series will be powered by the updated Intel Core Ultra 5 Series 3 processors, delivering 47 TOPS of NPU performance for all your AI tasks. On the other hand, the Vivobook S14 and S16 will feature sleeker metallic designs and Ultra 7 Series processors with up to 49 TOPS of NPU performance. Battery life for these is rated for 29 hours.