A nearly undetectable LLM attack needs only a handful of poisoned samples
Prompt engineering has become a standard part of how large language models are deployed in production, and it introduces an attack surface most organizations have not yet addressed. Researchers have de… #hackernews#llm#news
I have to suppress laughter and jeers every time someone in the office sounds desperate to implement AI into their jobs. Even had one guy today say that he's put simple tasks into Gemini and "it gets close!" lol you aren't going to design an entire house with AI you dorks. I work with babies.
“ ‘The use of free AI tools (e.g. ChatGPT, Claude, Gemini) for clinical purposes is strictly prohibited due to data security, privacy & accountability concerns .. also not allowed to use AI tools to draft notes & then transcribing it to handwritten’ “ www.rnz.co.nz/news/nationa...
“Rather than isolated cases of account misuse, the data points to a recurring pattern in which access to AI platforms is repeatedly advertised and redistributed through resale-style listings.” www.bleepingcomputer.com/news/securit...@bleepingcomputer.com
The latest update for #Snyk includes "The 5 Principles of Snyk's Developer Experience" and "From Discovery to Defense: Why #AI#RedTeaming Is the Next Step After AI-SPM".
- Why Anthropic discloses *literally nothing* on its energy and emissions impacts
- Why they're urging the US gov't to set a *target* (!!) for building fossil gas
In shutting down Sora—its app for AI-generated videos—OpenAI seems to “finally be learning that slop is not a business strategy,” @lilashroff.bsky.social argues:
Why ExecProtect matters long after an employee leaves ExecProtect is built to stop one of the most simple , but dangerous, email threats in healthcare: display name spoofing. Instead of relying on ...
MarkTechPost - Article
"NVIDIA Releases Nemotron-Cascade 2: An Open 30B MoE with 3B Active Parameters, Delivering Better Reasoning and Strong Agentic Capabilities"...
The latest update for #LevelBlue includes "#Cybersecurity in Hospitality: Defending a Highly Distributed Enterprise" and "LevelBlue and SentinelOne: Advancing Integrated, Intelligence‑Driven #SecurityOperations".
Source: Thrive Holdings has landed $1B in commitments for a new fundraising deal and is considering raising at least another $1B after strong investor interest (Bloomberg)
#WatchGuard announces new WatchGuard NDR for Firebox, Managed NDR and Total NDR offerings for organisations to easily deploy and scale AI-powered #threatdetection that uncovers, investigates and contains malicious activity.
The latest update for #CatoNetworks includes "Securing #AgenticAI: Why Visibility, Behavior, and Guardrails Matter" and "WebPromptTrap – New Indirect Prompt Injection Vulnerability in BrowserOS".
⚠️ ALERT: New scam using
CAPTCHAs, commonly used to verify that a user is human, often by asking them to click images or check a box. But in this scam, the page prompts users to follow a series of keyboard steps to continue.
The latest update for #CyCognito includes "Citrix NetScaler ADC and Gateway Vulnerabilities (CVE-2026-3055 & CVE-2026-4368)" and "Emerging Threat: Ubiquiti UniFi Network Application Path Traversal (CVE-2026-22557)".
AI SOC vendors are selling a future that production deployments haven’t reached yet
Vendors selling AI-powered security operations platforms have built their pitches around a consistent set of promises: autonomous threat investigation, dramatic reductions in analyst workload, and… #hackernews#news
The latest update for #CrowdStrike includes "How Charlotte AI AgentWorks Fuels Security's Agentic Ecosystem" and "Falcon #DataSecurity Secures Data Wherever It Lives and Moves".
The Law is already Established; Work for Hire Actually it's extremely likely. OpenAI itself cries foul and tells us that Deepseek “stole their IP” when it trained their models on OpenAI dat...
You can now transfer your chats and personal information from other chatbots directly into Gemini
Google is launching "switching tools" that allow users to transfer their chat history and personal information from other chatbots directly into Gemini. The feature aims to make…
The latest update for #WatchGuard includes "30 Years Driving Detection and Response in Hybrid Environments" and "SMB #Cybersecurity Spending Rises: #ZeroTrust & Secure Access Now Essential".
"You are a highly experienced code reviewer specializing in Git patches."
This is the first sentence of the default AI prompt on googlesource-dot-com. What does being a specialist in "git patches" have to do with general code review? wtf.
Anthropic won a court order blocking a Trump administration ban on government use of the company’s AI technology, after the Claude chatbot maker argued the move could cost it billions in lost revenue
What happens when AI starts checking mathematicians’ work – Mathematicians could soon use computers to verify proofs quickly and rigorously. Such a tool could help experts grapple with the accelerating pace and volume of mathematical research. A start-up call... https://tinyurl.com/25f5qo7y#UsingAI
…Coruna and DarkSword are two sets of advanced hacking toolkits that each contain a range of exploits capable of breaking into iPhones and iPads and stealing a person’s data, such as their messages, browser data, location history, and cryptocurrency….
The latest update for #Detectify includes "Introducing #GraphQL Support for API Scanning" and "Introducing IP Range Scanning: continuous Surface #Monitoring for your entire network".
The latest update for #ForwardNetworks includes "Why Your #AI Workflow Should Never Depend on a Single Model" and "Your Security Vision Has a Network Blind Spot".
Gateway Security Won’t Be Enough for MCP-Powered AI
As AI agents connect to enterprise tools via MCP, gateway-based security may fail. Here’s why policy enforcement must move to the MCP server. #hackernews#news
The latest update for #Foresiet includes "How Corporate Email Accounts Appear on Dark Web Markets (2026 India Edition)" and "Why 2026 is the Year of Proactive Cyber #ThreatIntelligence".
A federal judge on Thursday temporarily stopped the Department of Defense from labeling Anthropic as a security risk, in a reprieve for the artificial intelligence start-up and its work with the federal government.
“It's not only me and other researchers saying this, it's the lab CEOs themselves that [say] the risk is real.”
— Michaël Trazzi, an organizer with Stop the AI Race
At RSAC Conference, it was a privilege to moderate this panel on the future of the CVE program. Coverage of the discussion here —> www.nextgov.com/cybersecurit...
OpenAI and Anthropic before the IPO: Different balance sheets make comparison difficult
OpenAI and Anthropic are both experiencing rapid growth, but their revenue figures cannot be easily compared due to different accounting methods for cloud partnerships. The companies…
LLM Internals for Everyone: From Prompt to Answer What really happens when you hit “Enter” in ChatGPT, Gemini, Copilot, or Grok? Continue reading on Medium »
In response to a motion from Anthropic challenging the U.S. Government’s designation of the company as a supply chain risk, a federal judge issued a preliminary injunction in @AnthropicAI v. @DeptofWar.
💻 Federal judge grants Anthropic preliminary injunction vs Pentagon's blacklist, calling it punitive retaliation for refusing military use of Claude AI.[1][2] First US firm hit with supply chain risk...
The latest update for #Spike includes "How to route incidents based on what their payload says" and "How to set up Incident Alert Routing rules effectively".
Infiniti Stealer targets macOS using a fake Cloudflare page to trick users into running a Bash dropper. It steals browser creds, Keychain data, wallets, and dev secrets, exfiltrating via HTTP and notifying via Telegram. #InfinitiStealer#MacOSMalware
The latest update for #BitSight includes "The 2026 Ratings Algorithm Update: Strengthening Accuracy and Stability" and "Major Security Event: #SupplyChain Compromise in LiteLLM Versions 1.82.7 and 1.82.8".
Apple is abandoning its exclusive deal with OpenAI and will now allow other AI services to integrate with Siri and Apple Intelligence if their app is installed.
📰 Google Launches Search Live in Japan (2026): AI-Powered Voice & Visual Search Now Live
Google has launched Search Live in Japan, introducing AI-powered real-time voice and visual search powered by Gemini 3.1 Flash Live. The feature enables emotionally intelligent respo...
The latest update for #Veracode includes "Prioritize, Protect, Prove: A Roadmap for Application Security Transformation" and "Spring 2026 #GenAI Code Security Update: Despite Claims, AI Models Are Still Failing Security".
NY-based Blossom Health, which makes an "AI copilot" to augment psychiatrists' clinical decisions and automate office tasks, raised $20M in seed and Series A (Lily Mae Lazarus/Fortune)
The latest update for #Acronis includes "How to protect yourself from ransomware" and "#MSP#cybersecurity: how to choose a managed service provider that takes security seriously".
Cloudflare Workers and Tunnels are being exploited to host AiTM phishing pages and deliver malware like Xeno RAT and XWorm RAT, leveraging trusted domains to evade defenses and detection. #CredentialTheft#CloudAbuse#MalwareDelivery
4 Tactics for Shipping Faster Without Losing Software Quality
Speed vs quality is a false tradeoff. As systems scale, context becomes the real bottleneck. Top teams solve this by integrating context, building world models, using code simulations, and validating at the PR level. T… #hackernews#news
"Roose explains that both Meta and OpenAI have internal leaderboards that show how many tokens you’ve used, with one software engineer in Stockholm spending “more than his salary in tokens,” though Roose adds that his company pays for them"
Wait, so Anthropic calculates its annualized revenue by multiplying the *current subscribed users that day* by 12 plus its last four weeks of API by 13? Why the difference? Feels like you could game that quite easily www.theinformation.com/newsletters/...
Steno, which offers an AI-powered case transcript analysis tool for legal professionals, raised a $49M Series C led by Savano Capital Partners (Kyt Dotson/SiliconANGLE)
"As well as referring to Mythos, the draft blog post also discussed a new tier of AI models...Anthropic says: “’Capybara’ is a new name for a new tier of model..."
Capybara and Mythos appear to refer to the same underlying model."
In 2026, innovative AI companies like Google, Anthropic, Abridge, World Labs, and Mithril are leading the way, showcasing groundbreaking advancements in artificial intelligence and reshaping industries.
✨ The next 3D-BioInfo webinar will focus on deep learning and #3DBioInfo. Join the webinar to hear about deep learning for structured-based drug design and protein-ligand cofolding methods.
🗓️15 April 2026 | 🕔 17:00-18:00 CEST
Register now: https://loom.ly/aOcJfJE
Data professionals can now use AI tools for more than just generating code. Yu Dong takes a close look at how you might use Codex and MCP to connect Google Drive, GitHub, BigQuery, and analysis in one functional workflow.
Gemini 3.1 Flash Live makes AI audio sound more natural and fluid
Google is making its Gemini voice model more fluid, natural and precise while lowering latency and improving precision with the launch of Gemini 3.1 Flash Live. Gemini 3.1 Flash Live powers the recently announced Search Live, as…
Washington just signed a law requiring large AI platforms to watermark AI-generated content and give users tools to detect it. The law covers platforms with 1M+ monthly users — the state is home to Microsoft, Amazon, and Anthropic.
Gemini takes on ChatGPT and Claude with a major switching upgrade Easily import your memories and preferences Google has been rapidly improving Gemini, rolling out useful new features almost every ...
Cyber criminals continue to evolve not only how they can attack us, but also how they sell data and enable further cyber crime. New marketplace to sell stolen data targeting specific buyer demands.
#Anthropic is shifting the AI landscape from conversation to execution. The new "computer use" capability allows Claude to interact with any desktop application, navigating interfaces and performing multi-step workflows autonomously.
AI leap soon? Morgan Stanley says a big jump is likely in H1 2026 as compute soars. 10x more compute can 2x model intelligence. GPT 5.4 scores 83.0 percent on GDPVal with a 1 million token window. Gemini 3.1 hits 90 percent on IMO ProofBench. Benchmarks are not life.
Sources: Moonshot AI may scrap its Cayman structure for a China or Hong Kong entity to prepare for a Hong Kong IPO and plans to raise funding at ~$18B valuation (Wall Street Journal)
An extraordinary cybersecurity risk has spread across the national security community. It did not arrive through a breach, but through the tools everyone is already using.
New practice at OpenAI: instead of debating trade-offs in a design doc, they spin up multiple implementations in parallel and pick the one that works best. When prototyping costs near-zero, "let's just try all three" beats "let's discuss which one." www.youtube.com/watch?v=Bo6G...
The Energy Sector Isn’t Ready for Ransomware—and 2025 Proved It The global surge in energy sector ransomware attacks intensified throughout 2025, exposing deep vulnerabilities in critical inf...
A US federal judge suspended sanctions imposed by the Trump administration on Anthropic, saying the measures likely violated the law in blacklisting the AI company for expressing unease about the Pentagon's use of its technology.
A federal judge in San Francisco blocked a Pentagon order Thursday labeling the artificial intelligence company Anthropic a national security risk, saying officials had likely violated the law and retaliated against the firm for speaking publicly about how it wanted its technology to be used.
"A two-day selloff in memory-chip stocks is revealing a split in the artificial intelligence trade, as Google touts a breakthrough that analysts say may curb demand for certain types of storage while leaving others largely unscathed."
Brit lawmaker targeted by AI deepfake fails to get answers from US Big Tech
Appearing before Parliament, Meta, Google and X struggle to explain how fake political video circulated for so long
A member of the UK Parliament's lower house who was the victim of a deepfake AI c… #hackernews#meta#news
Anthropic's Mythos model introduces unprecedented cybersecurity risks in internal testing. As AI capabilities scale faster than defensive frameworks, the industry faces a critical gap. New threat models will be required. #AI#Securi…
Join the team driving Zero-Impact Breach Prevention.
At Mitiga, every role connects to a clear mission: stopping attackers before they cause impact. If you want to advance your career while shaping the future of cybersecurity, Mitiga is where you can do it.
When Verified Source Lies I deployed a staking vault on Sepolia and got it verified on Etherscan with a green checkmark. The source code contains a storage write that does not exist in the compiled...
if you work at anthropic your exponential backoff's first 5 tries are way too fast, you want to start at 1 second and then expo with double. idk how you're firing 5 retries in under three seconds that's nuts
“The researchers … found 1,748 valid credentials belonging to organizations including multinational corporations, critical infrastructure entities, and government agencies. The keys provide access to services like AWS, GitHub, Stripe, and OpenAI.”
Judge on the Anthropic case. "Nothing in the governing statute supports the Orwellian notion that an American company may be branded a potential adversary and saboteur of the U.S. for expressing disagreement with the government" www.nytimes.com/2026/03/26/t...
Codex now supports plugins. Integrations from Slack, Figma, Notion, Gmail, and more are available in the official Plugin Directory. #OpenAI#Codex#Plugins
After Funding Jolt, EU Moves to Back the CVE Vulnerability System The European Union is stepping forward to reinforce what many experts describe as a bedrock cyber vulnerability tracking system, ...
Windows Error Reporting Flaw Lets Attackers Gain SYSTEM Privileges Microsoft has patched a critical Windows Error Reporting (WER) vulnerability that allowed attackers to escalate privileges and gai...
A US judge temporarily blocks the Pentagon's blacklisting of Anthropic, halting military restrictions in the latest legal battle over AI safety. www.jpost.com/internationa...
Great, great story but when you're writing about AI companies not covering their tracks well at least strip out the UTM tracking from your links showing you're looking for past reporting on your own site using ChatGPT fortune.com/2026/03/26/a...
Critical Citrix NetScaler and Gateway Vulnerabilities Let Remote Attackers Leak Sensitive Information Cloud Software Group has issued a critical security bulletin detailing two newly discovered vul...
💻 **Anthropic secures federal injunction vs Trump admin, forcing reversal of Defense Dept's "supply chain risk" label and agency bans over AI usage disputes.** (118 chars)[1]
ActionPower targeted in a ransomware attack by threat actor crypto24, involving file encryption and possible data theft. Incident traced to South Korea, discovered March 27, 2026. #Crypto24#RansomwareAttack#SouthKorea
CirrusHQ makes sales appointment amid shift to customer resilience
SCOTLAND-BASED cloud specialist CirrusHQ has appointed Matt Smith as head of sales, bolstering its senior leadership team as the business targets growing demand for cyber resilience and cloud-based security support across the UK. …
'We're going to search your house. We don't have a warrant, but we paid your landlord $100 to give us a spare key. So now we're searching your house without a warrant,' Laperruque www.npr.org/2026/03/25/n... from @npr.org
Running iOS 18? Update now, as there's an HTML and JavaScript exploit that exfiltrates contacts, messages, call history, and keychain data from the previous version of the OS that is so secure you're forbidden from running a browser that's not made by the same vendor. techcrunch.com/2026/03/23/s...
Due to software vulnerability in products of US manufacturer PTC, the state criminal police in #Germany visited admins at home and rang them out of bed: digit.site36.net/2026/03/27/h...
Booz Allen just dropped Vellox — an AI-native cyber product suite built by a defense contractor with decades of state-sponsored threat intel.
When attacks break out in seconds, human-speed defenses won't cut it. www.shashi.co/2026/03/booz... #CyberSecurity#AI#RSAC2026
Researchers release tool to detect stealthy BPFDoor implants in critical infrastructure networks
Telecommunications providers around the world have been dealing with the burrowing efforts of the China-linked APTs for many years now. To help them identify hard-to-detect implants u… #hackernews#news
Coruna iOS exploit framework linked to Triangulation attacks
The Coruna exploit kit is an evolution of the framework used in the Operation Triangulation espionage campaign, which in 2023 targeted iPhones via zero-click iMessage exploits. [...] #hackernews#news
The company is taking new steps to stop #AI impersonation — but across the internet, the problem continues to grow. PLUS: Anthropic in court, and Meta loses in New Mexico
Google Gemini fires major AI shot! Memory Lane imports chat history from ChatGPT Claude. Bye lockin hello usercentric AI! #GoogleGemini #AIWar #MemoryLane #ChatImport #UserCentric #TechNews
💻 OpenAI rolled out ChatGPT ads in the US on free/Go tiers after testing. User tested 500 Qs to reveal most frequent sponsored promos tied to prompts.[2][5] (124 chars)
Sumo — Shellshock on a CGI Script and Dirty COW on a 2012 Kernel | OffSec PG Sumo is a simple machine that combines two well-known CVEs into one exploit chain. The box is running an ancient Apach...
Apple takes a step towards Android by opening up Siri to third-party AI in IOS 27 Your iPhone is about to get way smarter as Apple opens the door to rival AI. Credit: Ryan Haines / Android Authorit...
TikTok for Business accounts targeted in phishing campaign — here's how to stay safe TikTok for Business users warned not to click any shady links, especially those recently registered throug...
Google Gemini now allows users to import memory, preferences, personal context, and chat history from other AI apps like ChatGPT and Claude, ensuring a seamless transition without losing key information or prior conversations. alternativeto.net/news/2026/3...
TikTok for Business accounts targeted in new phishing campaign
Threat actors are targeting TikTok for Business accounts in a phishing campaign that prevents security bots from analyzing malicious pages. [...] #hackernews#news
Data leak reveals Anthropic is testing 'Claude Mythos,' a powerful new AI model representing a "step change" in capabilities. Early access customers are currently trialing it. #AI #AI#News
In the coming weeks, OpenAI will begin expanding ads beyond the U.S., starting with pilots in Canada, Australia, and New Zealand. openai.com/index/testin...
"In January, a meeting between company executives and its council of advisers got heated, with one of the advisers cautioning that OpenAI could be in the process of developing a “sexy suicide coach,” The Wall Street Journal previously reported."
Bearlyfy, a pro-Ukrainian group, has hit 70+ Russian firms since 2025, evolving from LockBit 3 and Babuk to deploying custom GenieLocker ransomware and sophisticated tools for high-value extortion. #Bearlyfy#GenieLocker#Russia
OpenAI's Codex gets a plugin marketplace for Slack, Notion, Figma, and more
OpenAI is launching a plugin marketplace for Codex that integrates with workplace applications including Slack, Figma, Notion, Gmail, and Google Drive. The plugins will allow Codex to interact…
As it scales, it’s outpacing legacy defenses and reshaping security. To stay resilient, orgs must tackle 3 frontiers:
🔹 Data Poisoning
🔹 AI-Driven Phishing
🔹 Shadow Cloud Governance
“In a scathing 43-page ruling, Judge Rita F. Lin of the U.S. District Court for the Northern District of California said Anthropic would not be restricted from continuing with its federal contracts for now. The ruling is not a final decision ..” www.nytimes.com/2026/03/26/t...@nytimes.com
TeamPCP Hackers Target AI Developers with Malicious Code Injections A large-scale software supply chain attack linked to the threat actor group “TeamPCP” has triggered a high-priority alert fro...
Having a diversely of information space systems is healthy for all democratic instructions. I support Anthropic in this situation, however I also support an infinite number of other Ai offerings and protocols bridging trust networks.
John Martinis, 2025 Nobel Prize–winning physicist and former Google quantum lead, joins the White House science panel to help shape U.S. tech policy, highlighting focus on emerging fields like quantum computing.
Med News: #AI tools are changing how patients access health info.
From explaining medical terms to proving postoperative instructions to digitizing informed consent, experts explore how these technologies may serve as “clinician extenders.”
Claude Mythos: Leak spills details on Anthropic’s new AI model, its most powerful yet: A data leak has revealed that Anthropic is developing "Claude Mythos," its most powerful AI model to date, currently in early testing. Internal documents, accidentally exposed onlin… https://ranked.news/410562?u=b
CISA warns of a critical deserialization vulnerability (CVE-2026-4681) in PTC Windchill and FlexPLM that allows remote code execution without authentication. German police have issued physical alerts. #PTCFlaw#Germany#RemoteCode
LISTEN: Here’s where cyber and physical threats converge #LATEST#TECHNOLOGY
https://dawnnewsup.com/2026/03/27/technology/listen-heres-where-cyber-and-physical-threats-converge/
ChatGPT “Spud” : What We Know About OpenAI’s Next GPT AI Model Evolution OpenAI’s upcoming “Spud” model is poised to make a significant impact on the global economy, with its potential ...
Internal messages of concern about platform decisions were used to great effect by the plaintiffs' attorneys in the social media cases. As @masnick.com observes, this will lead companies to shut down internal dissent channels or internal reviews of risks. www.techdirt.com/2026/03/26/e...
ChatGPT erotic chatbot? OpenAI says no Concerns over harm, moderation, and AI relationships have pushed adult mode off the table Users looking to bring a little spice to the best AI phones , take n...
Unity plans to shut down its ironSource advertising network on April 30 and has begun the process of divesting Supersonic, its casual game publishing arm (Sophie Shulman/CTech)
How to build HIPAA-compliant AI agents in healthcare, BAA requirement, PHI data rules, FHIR scoping and HITL checkpoints explained for CTOs and compliance teams #healthcareappdevelopment
Opening DIGIT’s Scot-Secure Summit, McDonald’s senior threat analyst Ellie Hallam argued businesses must adapt to survive the flood of AI-driven malware, deepfake propaganda, and state-sponsored threats blurring the lines between military and civilian infrastructure.
🤖 [AI] LEAKED DOCUMENTS FROM ANTHROPIC SHOW THAT A NEW GENERATION OF SUPER-STRONG MODELS, "CLAUDE MYTHOS," IS ALREADY IN TESTING WITH ANTHROPIC BELIEVING IT "POSES UNPRECEDENTED CYBERSECURITY RISKS," PER FOR...
VoidLink Rootkit Exploits eBPF and Kernel Modules For Stealth On Linux VoidLink is a sophisticated, cloud-native Linux malware framework that leverages a rare hybrid architecture to maintain deep s...
“The EU wants to help ‘build upon’ the foundation of the program & ‘the great work that has been done there,’ Hans de Vries, the chief cybersecurity and operational officer for the European Union Agency for Cybersecurity, or ENISA, said Thursday ..” www.nextgov.com/cybersecurit...
🤖 [AI] THE PENTAGON HAS BEEN BLOCKED BY A US COURT FROM PUNISHING ANTHROPIC OVER ITS REFUSAL TO ALLOW UNRESTRICTED USE OF ITS TECHNOLOGY IN WARFARE, PER FT ...
$AMZN "Garman explained ... there’s enough pent-up AI demand to keep Amazon’s cloud tools busy for the next five to 10 years even if the tech stops advancing." www.cnn.com/2026/03/27/t...
Federal judges stepping into AI contract disputes feels like we're entering uncharted territory. The legal system wasn't built for this pace of tech policy.
AI company Anthropic has inadvertently revealed details of an upcoming model release, an exclusive CEO event, and other internal data, including images and PDFs, in what appears to be a significant security lapse.
Opinion: The Nippon–OpenAI lawsuit highlights the potential legal and ethical implications of AI systems providing professional advice without proper oversight. Companies should adopt a layered strategy to protect against AI-generated legal documents and lawsuits.
Here’s this week’s Better Offline monologue. I talk about OpenAI killing Sora, and how the majority of announced data centers are barely getting built or vaporware, with only 3GW coming online in 2025, and only 5GW under construction.
A Group-IB report published on March 25 highlights the rising threat of financial fraud linked to cloud phone technology. These remote-access Android devices mimic legitimate smartphones, complicating fraud detection. In the UK, Authorized Push Payment fraud losses reached £485.
anthropic apparently declines to allow Claude in Chrome to operate on the NYT website, which includes NYT Cooking, which is putting a real damper on my meal-planning automation trial
Web shells are malicious scripts giving attackers persistent access to compromised servers for command execution, data theft, and launching attacks. Exploits include SQLi, RFI, XSS, and stolen credentials. #WebShells#ServerSecurity#InfoSec
A dangerous piece of malware is public and easy to implement. You can protect yourself simply by updating your iOS device: leafandcore.com/2026/03/27/y...
Apple says no one using Lockdown Mode has been hacked with spyware The tech giant's claim that it has not seen any successful spyware attacks targeting Apple devices with Lockdown Mode enabled ...
More than half of the crimes registered in the city of Tashkent in 2025 were cybercrimes. With the expansion of digital services and financial transactions conducted online, criminals are following the money.
The negotiations between the United States Department of Defense and Anthropic have highlighted who gets to make choices about #AI in the #military. Who should get to make these choices? spectrum.ieee.org/military-ai-...
As higher ed institutions pay tech companies millions to provide students and faculty access to custom AI-powered tools, some faculty in Colorado and California are pushing back. https://bit.ly/4bBIZHv
1/ 🤖 Anthropic, the "safety-first AI company," accidentally leaked details of a new model Fortune says poses "unprecedented cybersecurity risks." Irony is working overtime this week.
The internet is no longer mostly human: Automated traffic grows 8x faster than human traffic. AI-driven traffic is now the fastest-growing category online and for the first time, AI isn't just browsing. It's buying... 📈 www.humansecurity.com/learn/resour...#AIAgents#GEO#AICommerce#FutureOfWeb
Sources: Aetherflux, founded by Robinhood co-founder Baiju Bhatt to develop orbital data centers, aims to raise $250M to $300M in a Series B at a $2B valuation (Wall Street Journal)
Intelligent, frank, and thought-provoking content can be hard to find these days. If you enjoy tickling your gray cells with good podcasts, check out SoulTech with Cynthia. Learn more w/ trailer.
"According to a recent investigation by cybersecurity firm Gen Threat Labs, more than 18,000 OpenClaw instances are already exposed to internet attacks, and almost 15 percent of them contain malicious instructions." #edtech#ILoveEdTech#ImFutureReady#elearning#AIEdu
"A draft blog post that was available in an unsecured and publicly-searchable data store prior to Thursday evening said the new model is called “Claude Mythos” and that the company believes it poses unprecedented cybersecurity risks."
Elastic Security Labs uncovered BRUSHWORM, a modular backdoor spreading via USB and stealing files, and BRUSHLOGGER, an XOR-encoded DLL side-loading keylogger targeting a South Asian financial institution. #MalwareAnalysis#SouthAsia#Backdoor
Apple's latest announcements are here! From the sleek MacBook Neo to the immersive AirPods Max 2 and the innovative iPhone 17e, there's something for everyone. Discover the future of tech with these excit…
An Amsterdam court issued an injunction on Thursday prohibiting Elon Musk’s xAI from allowing its chatbot Grok to keep creating AI CSAM & other non-consensual intimate images. (Basically xAI has to disable Grok's nudify features or face daily fines)
Figure AI Founder Bets on ‘Family’ of AI Devices with New Venture Hark Brett Adcock launches Hark to build AI devices as OpenAI, Apple, and Meta push into hardware, raising questions about the ...
Strengthening cybersecurity in education through private sector partnership As schools are increasingly targeted by cybercriminals, private sector investment equips them with tools to mitigate atta...
“What made the attack impressive was the next move: The scammers… contacted Apple Support themselves, pretending to be me, and opened a real case…” ma.tt/2026/03/gone...
“In preparing to release Claude Capybara, we want to act with extra caution and understand the risks it poses—even beyond what we learn in our own testing. In particular, we want to understand the model’s potential near-term risks in the realm of cybersecurity,”
This week in cybersecurity by the numbers: Hackers hit a benefits provider, an anime streaming service, and a crime tip platform—exposing millions of sensitive records.
🇪🇺🇯🇵 The EU & Japan held their 9th Digital Policy Dialogue in Tokyo, advancing cooperation on #semiconductors, #AI, #quantum, #cybersecurity & #5G/6G — all core to INSTAR's work on international digital standards with like-minded partners.
Meta agrees to fund Entergy Louisiana's new energy infrastructure for its Louisiana data center, including seven natural gas power plants (Nicholas G. Miller/Wall Street Journal)
Keitaro’s all-in-one tracker, cloaker, and TDS tools are exploited by threat actors to rapidly scale scams, phishing, malvertising, and malware distribution including DonutLoader campaigns and wallet-drainers. #AdtechAbuse#Malvertising
VCs are betting billions on AI’s next wave, so why is OpenAI killing Sora?
An 82-year-old Kentucky woman rejected a $26 million offer from an AI company seeking to build a data center on her land. The incident highlights growing resistance from local communities as AI…
RevOps AI: Build Your Entire Sales Team on Notion with Gemini and MCP
How I built a full Revenue Operations platform using Notion as the database, Gemini 2.5 Flash as the AI agent, and the Model Context Protocol to wire 22 Notio…
Apple’s Lockdown Mode feature is the best defense we have against spyware on iOS, macOS, watchOS, and iPadOS. Apple launched the feature four years ago and has not yet seen a device with Lockdown Mode on be compromised. techcrunch.com/2026/03/27/a...
What is #MCP, and why does it matter for #WordPress? We explain how the Model Context Protocol connects AI tools to your site and what it could enable going forward → speckyboy.com/what-is-mcp/
Cybersecurity stocks slumped on Friday on a report that Anthropic is testing a powerful new artificial intelligence model called Mythos that presents potential security risks.
The writers getting falsely accused of using A.I., OpenAI is shutting down Sora, the tech bros obsessed with "taste" (people always want what they can't have), and more of the best writing and books links from the past week!
Telnyx Python SDK versions 4.87.1 and 4.87.2 on PyPI were compromised with credential-stealing malware hidden in src/telnyx/_cl. PyPI quarantined the releases; users should revert to 4.87.0 or earlier. #SupplyChain#PythonSDK#USA
A preliminary injunction in the Anthropic case means that the underlying issues extend beyond just a dispute over procurement policy or the appropriate redlines for government use of AI technology, says the Cato Institute’s Jennifer Huddleston.
The future of online shopping just got a boost! A new tool has been launched to verify the humans behind AI shopping agents, enhancing trust and transparency in e-commerce. Discover how this innovatio…
Integrity360 CTO Richard Ford talks to us about the meaning behind the organisation’s cybersecurity conference Security First, which recently came to Dublin’s Aviva Stadium. See the full interview 🔗 vist.ly/4wdri
