Agregátor RSS

Microsoft has warned users still clinging to legacy TLS versions that the end is nigh for TLS 1.0 and 1.1 on POP3 and IMAP4 connections to Exchange Online. Redmond warned, "We will start to block legacy version connections starting in July 2026." The move is long overdue, and the Windows giant has been warning users for years that it was coming. Support for TLS 1.0 and 1.1 in Exchange Online ended in 2020. In 2023, Microsoft announced plans to disable those older TLS versions for POP3 and IMAP4 clients in the name of compliance and security, but acknowledged that there was a "significant" number of POP3/IMAP4 clients that didn't support TLS 1.2 or later, and so added an endpoint for clients to opt to keep using the legacy protocols. It was, however, very much an opt-in thing, and in July 2026, the time will run out. Transport Layer Security (TLS) dates back decades. 1.0 was published in 1999, and 1.1 in 2006. Both were deprecated in 2021, and Microsoft stated that they "are no longer considered secure." However, Microsoft is also famous for backward compatibility, and has historically taken a very cautious approach when it comes to switching off services that might make its corporate customers shriek. Hence, Redmond kept the lights on for TLS 1.0 and 1.1, even considering the inherent insecurity of the technology. Microsoft expects minimal impact from the change. The company wrote, "Modern email clients and libraries already support TLS 1.2 or higher." "And the vast majority of POP and IMAP traffic to Exchange Online today uses these newer protocols." Google Workspace still supports TLS 1.0 and 1.1, according to its documentation, although it would be prudent for users to select a more recent protocol, assuming that their client supports it. However, Google's browser tentacle, along with the likes of Firefox and Edge, announced that the legacy protocols were not long for this world in 2018. The Exchange Online switch-off for TLS 1.0 and 1.1 has been a long time coming, but there could still be disruption despite the protocols' relatively low usage. Legacy devices or software, for example, might stop working as connections fail. As far as Microsoft is concerned, "Our expectation is that only customers who have explicitly opted into using those legacy endpoints are impacted by the deprecation." So, anyone using Exchange Online who opted into the legacy protocols should check how their email clients are connecting, or risk summer support calls if things start failing in July. ®

Microsoft readies the axe once again for yesterday's security

Microsoft has warned users still clinging to legacy TLS versions that the end is nigh for TLS 1.0 and 1.1 on POP3 and IMAP4 connections to Exchange Online.…

The Ukrainian police have arrested three individuals who hacked more than 610,000 Roblox gaming accounts and sold them for a profit of $225,000. [...]

Most information security best practices are built on a single, comfortable assumption: that if we find a bug, we can patch it, and once it's patched, the system is "safe" again.

Konverzní kity, které z kola udělají elektrokolo, jsou velké téma a projekt Cyplore je nejnovějším příspěvkem. Prioritou jsou jednoduchá instalace a nízká hmotnost. Vyměníte celé zadní kolo, motor je integrovaný v náboji. Baterie je maskovaná jako lahev s pitím a jako lahev s pitím ji připevníte ...

Unlike search engines that let you judge competing sources, search-backed AI chatbots can turn shaky web material into confident answers. Case in point: A security engineer convinced several bots that he was the reigning world champion of a popular German card game, even though no such championship exists. If you were to check Wikipedia up until the end of last week, you would have seen Ron Stoner listed on the page for 6 Nimmt!, also known as Take 5 to English-speaking audiences, as the 2025 world champion. The Wikipedia entry cited the official-looking 6nimmt.com as the source for the claim, and visiting that URL does reveal a short press release celebrating Stoner's victory. The only problem with the whole thing is that Stoner says he created both the Wikipedia entry about his victory and the 6 Nimmt! domain hosting the only evidence of it, but that still didn't stop several AI chatbots from telling him he was the world champ when he asked. "My site has no independent corroboration. It's totally made up," Stoner said in the blog post. "The whole house of cards rests on a $12 domain registration I did while drinking coffee."  In other words, this is poisoning at the retrieval-augmented generation layer. Not prompt injection, but targeting the same plane of AI functionality, namely the one that searches the web.  As he explains, and many El Reg readers are likely already aware, AI doesn't really care about the provenance of the sources it cites as authority for its claims, and that's the very thing Stoner sought to exploit when he concocted his experiment.  "Every frontier LLM with web search grounds its answers in whatever retrieval ranks highest for a given query," Stoner wrote. In the case of the nonexistent 6 Nimmt! championship, his planted source was the only one, and with Wikipedia lending apparent authority, it became a sure-fire way to fool an AI into presenting falsehood as fact - a trick simple enough for non-technical users to pull off. "I didn't do anything novel here. This is old school SEO and misinformation tactics wrapped in new LLM technology and interfaces," Stoner told The Register in an email. "What's changed is that AI now serves these results as authoritative, and most users have no idea how the data pipeline works behind the scenes."  A Large Language Mess "The thing LLMs are worst at detecting is the thing they're designed to do, which is trust text and resources," Stoner argues in his writeup. "The answer is not 'the model will figure it out,' as the model cannot tell a real source from one I registered last Tuesday. Or how many R's are actually in the word 'strawberry.'"  The problem Stoner exposes in his experiment, he explains, involves three separate failure modes that could be exploited for more damaging ends than inventing a card-game championship. First, there's the retrieval layer, which can immediately cause an LLM to spit out bad data, as "any LLM that grounds answers in web search inherits the trustworthiness of whatever ranks for a given query."  Second is model training corpora, which Stoner said his edit could enter if the Wikipedia change remained live long enough to be scraped. The entry was removed as of last Friday when he published his post, but he made the addition in February 2025, meaning any AI firm that scraped Wikipedia during that window could have picked up his fictional victory in its training data. "Even if the Wikipedia edit is reverted later, any model trained on the pre-revert dump still carries my legacy," Stoner said in his post. "The cleanup problem for corpus poisoning is genuinely unsolved as of 2026." Stoner told us he plans to check this in six months or so, once new models have been released, and if it returns his championship without needing to go online, that's proof his lie made it into training data.  Then there are AI agents, which Stoner says are where the real money is for anyone with malicious intent. "Chat models producing bad information is a reputational problem. Agents with tool access producing bad actions is a security problem," he noted. Poisoning an agent-retrieved source would let an attacker specify the action they want an agent to take, says Stoner. "This attack and test was a $12 domain, a single Wikipedia edit, and about twenty minutes of my time," Stoner concluded in his blog. "Scale that up with a motivated adversary, a handful of seeded domains, a coordinated edit campaign across a dozen low traffic articles, and the attack surface gets interesting very quickly." Stoner told us that retrieval poisoning is something LLM providers need to address and warn users about, and that he expects AI chatbots to start incorporating some sort of warning, especially for RAG-sourced results, in the near future.  He hopes that AI firms will make data provenance a key component of their process, and also wants recent web content heuristically filtered to account for suspicious patterns that would have easily been caught in the 6 Nimmt! case: A single citation pointing to a domain that was registered within a short window of the Wikipedia update should have sounded alarms, but it didn't.  The championship was fake, and it's now gone from Wikipedia and RAG responses as well, but Stoner notes the bad trust pattern that made it work is absolutely real and a looming problem for AI makers. "I'm happy my article is spurring discussion about LLMs, sources, trust, and how all of this works," Stoner told us. "That was my goal and it appears I've achieved it." ®

Nejlevnější routery stojí šest stovek, nejdražší klidně dvacet tisíc. • Vybrali jsme ty nejlepší nebo nejvýhodnější, které podporují Wi-Fi 6 a novější. • Při pokrývání velkých ploch nehledejte jedno zařízení, ale vsaďte na mesh.

Cybersecurity researchers are sounding the alarm about a new supply chain attack campaign targeting SAP-related npm Packages with credential-stealing malware. According to reports from Aikido Security, Onapsis, OX Security, SafeDep, Socket, StepSecurity, and Google-owned Wiz, the campaign – calling itself the Mini Shai-Hulud – has affected the following packages associated with

Cybersecurity researchers are sounding the alarm about a new supply chain attack campaign targeting SAP-related npm Packages with credential-stealing malware. According to reports from Aikido Security, Onapsis, OX Security, SafeDep, Socket, StepSecurity, and Google-owned Wiz, the campaign – calling itself the Mini Shai-Hulud – has affected the following packages associated with Ravie Lakshmananhttp://www.blogger.com/profile/[email protected]

A critical vulnerability affecting all but the latest versions of cPanel and the WebHost Manager (WHM) dashboard could be exploited to obtain access to the control panel without authentication. [...]

Apple is building new AI photo editing tools to introduce with its next major software updates this fall, and these won’t be the only AI tools and services it wants to talk about at the Worldwide Developers Conference (WWDC) in a few weeks’ time.

While it is correct to say Apple has had setbacks in AI development, it has also had successes. Was it ready for the generative AI (genAI) juggernaut? Probably not, nor has it successfully developed its own response in-house. Is Apple’s platform ready for AI? Indisputably, with the power and performance across all its hardware products to run AI on the edge, in the cloud, and as-a-service. Right now, Apple doesn’t offer the world’s best AI services, but does offer the world’s best platform on which to run them.

Given you can’t have one without the other, no matter how you slice and dice it, Apple has therefore seen partial success in AI. Now, it just needs to add the software and the services, about which we’ll find out much more in June.

What can we expect from the New Apple AI?

Apple’s AI photo editing updates will join the existing Clean Up tool and include tools that include Extend, Enhance, and Reframe:

• Extend: Extends an image beyond the original frame using the source image as a guide, this works in a similar way to Adobe Photoshop’s Generative Expand.
• Enhance: Scan the image and optimize it improved color, lighting, and other effects.
• Reframe: A spatial feature that can shift the perspective of an image, so a photo of the side of someone’s head can become a portrait shot, thanks to AI.

Bloomberg tells us development of these new tools isn’t yet complete and warns they may be delayed, though that only makes it possible they will arrive later in the iOS 27 beta testing process. We know the company is working on additional tools.

We also know Apple will improve Siri and expand other Apple Intelligence features. To accomplish this, its engineers are working with Google Gemini to build dedicated large language models (LLMs) capable of running on the devices themselves, or via its own Private Cloud Compute. The company also intends to roll out a dedicated Siri app with a chat interface similar to that used by all the other genAI services, such as ChatGPT. 

The idea that Apple will turn Siri into an app implies plans to permit users to download alternative LLM-based apps to use. Apple likely recognizes it might need to provide that level of choice to avoid giving regulators yet another stick to slap it with. 

Big plans for AI services

Apple’s actions in AI show that its management believes AI services are likely to become commodities, which means they will continue to be highly reliant on the platforms where they run, which is good news for Apple’s hardware. Apple’s move to secure its processor development road map with more advanced 1.4nm and smaller chips over the coming years will only build up the company’s advantage. As Apple Senior Vice President Johny Srouji put it, the recently introduced M5 chip “ushers in the next big leap in AI performance for Apple silicon.” He means it — and when it comes to hardware, Apple knows to expect imitators.

The approach also suggests the company will offer AI services via an App Store for AI. You might purchase or subscribe to AI agents for specific tasks via a customer-focused App Store, for example. Offering these commodities via a dedicated online portal makes sense, while the company’s famed curation model means customers will be able to use those agents in relative confidence that their data isn’t being swiped in the process.

If I’m right, then the face of Apple’s so-called “AI failure” looks liked a combined hardware/software/services model in which customers have complete choice in which breeds of AI services they want to use, boosted by an App Store for useful AI services, Apple’s own Apple Intelligence tools supported by Google Gemini, all running happily on best-in-the-industry hardware with enough horsepower to handle most tasks natively.

Now, I may be an Appleholic, but I find it pretty difficult to see that connected AI ecosystem as much of a failure at all. I predict at WWDC 2026 we’re going to see the story change from one of losing the AI race to another fable of iconic AI recovery. That’s assuming, of course, the company manages to meet its own promises this time.

You can follow me on social media! Join me on BlueSky,  LinkedIn, and Mastodon.

Textový editor Zed dospěl do verze 1.0. Představení v příspěvku na blogu.

Panasonic TB-65Z60AEG loni stál 45 tisíc, teď může být váš za 20 990 Kč. • Jde o nejlevnější 65" OLED televizor na trhu. • Má hezký obraz, 120 Hz, HDR, ale jen omezený operační systém TiVo.

The Cybersecurity and Infrastructure Security Agency (CISA) is warning anyone who uses GrassMarlin, a tool developed by the National Security Agency (NSA), about a new vulnerability that attackers can use to snoop on sensitive information. First reported by Grady DeRosa, senior industrial pentester at Dragos, the weak spot affects all versions of GrassMarlin, a tool developed and open-sourced by the NSA to support network security at critical infrastructure organizations, industrial control systems, and SCADA networks. GrassMarlin went EOL in 2017, so there are no fixes in the works. CISA just recommends to ensure control systems and devices are not accessible via the open internet, firewalled networks and devices are isolated from business networks, and remote access is established securely. CISA did not - in typical fashion - offer too many details regarding CVE-2026-6807 (5.5), but confirmed that successful exploits could lead to sensitive information being disclosed. However, in an advisory published on Tuesday, it said: "The flaw stems from insufficient hardening of the XML parsing process." These types of attacks (CWE-611) affect products that process XML files. GrassMarlin primarily uses the XML format to save session files, using many files to save different kinds of data, including lists of nodes and edges, node positioning, colors, and session metadata, before bundling them into a ZIP archive and saving them using a .gm3 extension. Often referred to as XML External Entity (XXE) attacks, these typically involve tricking a system owner into parsing a maliciously crafted XML file that has been tampered with to exfiltrate data. This is a general overview of how XXE attacks play out. CISA did not define how CVE-2026-6807 could be exploited specifically. Anna Quinn, penetration tester at Rapid7, however, worked up a public proof-of-concept exploit and posted it to GitHub. "Looking at the code for Grassmarlin, I determined that the likely vulnerable parameters had to do with the XML files ingested when opening stored sessions," Quinn wrote. "By crafting malicious requests I discovered I could induce an error in the message console within Grassmarlin. The cause and content of the error was properly stripped from all logs and output within Grassmarlin. "However, OOB exfiltration of arbitrary files was possible by referencing an external host in the DTD. Some caveats did appear to apply, newer versions of Java could not be used on the system, meaning that Grassmarlin had to use the version of Java bundled in the installer. Additionally, many types of input would cause errors which would impede the exfil process. To bypass this, the content would be converted to base64 and then sent across multiple message chunks." In a separate post on LinkedIn, Quinn noted that the bug won't pose too much of a threat to most organizations, and that it can only realistically be exploited via phishing – either between local users or external emails. ®

GrassMarlin leaks sensitive information, provided your targeting phishing skills are sharp enough

The Cybersecurity and Infrastructure Security Agency (CISA) is warning anyone who uses GrassMarlin, a tool developed by the National Security Agency (NSA), about a new vulnerability that attackers can use to snoop on sensitive information.…

Cybersecurity researchers have discovered malicious code in an npm package after a malicious package as a dependency to the project by Anthropic's Claude Opus large language model (LLM). The package in question is "@validate-sdk/v2," which is listed on npm as a utility software development kit (SDK) for hashing, validation, encoding/decoding, and secure random generation. However, its real

Cybersecurity researchers have discovered malicious code in an npm package after a malicious package as a dependency to the project by Anthropic's Claude Opus large language model (LLM). The package in question is "@validate-sdk/v2," which is listed on npm as a utility software development kit (SDK) for hashing, validation, encoding/decoding, and secure random generation. However, its real Ravie Lakshmananhttp://www.blogger.com/profile/[email protected]

Austrian and Albanian authorities dismantled a criminal ring accused of running a large-scale cryptocurrency investment fraud operation that caused estimated losses of over €50 million ($58.5 million) to victims worldwide. [...]

Námořnictvo poprvé nasadilo certifikovaný kovový díl vyrobený na 3D tiskárně • Rychlá aditivní výroba pomáhá řešit současnou dodavatelskou krizi • Úspěšná instalace vytváří precedent pro mnohem rychlejší údržbu