This is the first post in a series using BishopFox's BadPods, which features a variety of pod manifests with varying levels of misconfiguration, to investigate the protections included in different cloud providers' managed Kubernetes offerings. This series includes the most permissive pod manifest, which the author uses to trivially escape the pod via a mounted host path and then further escalate privileges within the account by stealing IAM credentials from IMDS.
BinaryAudit is an open-source benchmark from Quesma for evaluating AI agents' ability to detect backdoors in stripped binary executables using reverse engineering tools like Ghidra and Radare2. The benchmark tests models against backdoored versions of lighttpd, dnsmasq, Dropbear, and Sozu, with Claude Opus 4.6 leading at 49% detection but hampered by a 22% false positive rate across clean binaries. Results highlight that while AI-assisted binary analysis is becoming accessible to non-specialists, high false positive rates and inability to trace data flow through complex binaries make it unsuitable for production malware detection today.
Nucleus provides an exposure management orchestration platform that ingests data from 200+ security and asset tools, normalizes and correlates vulnerabilities with threat intel and business context to prioritize and remediate risk across environments.
Dotenv is a zero-dependency module that loads environment variables from a .env file into process.env, storing configuration in the environment and separating from code while following the Twelve-Factor App methodology.
The Kimwolf IoT botnet accidentally disrupted the I2P anonymity network after attempting to register 700,000 infected devices as nodes — a Sybil attack that overwhelmed the roughly 15,000-20,000 node network — while seeking takedown-resistant C2 infrastructure. The botmasters, who admitted to the disruption on Discord, have been experimenting with both I2P and Tor as fallback command-and-control channels after security companies and network operators intensified efforts to combat the botnet. I2P is currently operating at roughly half capacity as stability fixes roll out, while Kimwolf's overall numbers have reportedly dropped by more than 600,000 infected systems following internal operator disputes and operational errors.
Microsoft has shipped emergency fixes for several actively exploited Windows and Office zero-days that enable one-click remote code execution and SmartScreen bypasses across all supported Windows versions. Attackers can silently run high-privilege malware, enabling ransomware and espionage campaigns. Admins and end users are urged to patch immediately and treat malicious links and Office files as high risk.
Google's threat team details how North Korean group UNC2970 and others weaponize Gemini for OSINT-driven target profiling, tailored phishing, vulnerability analysis, and exploit debugging, blurring the line between research and reconnaissance. Attackers also use Gemini's API in HONESTCUE malware, AI-built COINBAIT phishing kits, and large-scale model extraction attempts that clone Gemini's behavior via high-volume prompt queries.
Law enforcement crippled Lumma Stealer's infrastructure in 2025, but the infostealer has rapidly rebuilt and is spreading again using persuasive “ClickFix” fake CAPTCHA prompts that trick users into running commands in Windows Terminal, loading CastleLoader in memory, then Lumma.
Google Cloud Security conducted a five-month security audit of Intel's Trust Domain Extensions and discovered five vulnerabilities.
Russia is escalating its crackdown on independent messaging platforms by attempting to fully block WhatsApp and throttling Telegram.
Social media continues to play a growing role in AI sourcing. Reddit dominates with 3-4% of all citations, 99% of which are individual threads rather than profiles. YouTube and LinkedIn reward authority, with 65% of YouTube citations pointing to channels and 47% of LinkedIn citations to personal profiles. Instagram favors content, as reels and posts make up 65% of citations versus 30% for profiles.
A Forrester poll in the US, UK, and Canada found that 83% would accept ads in AI chatbot free tiers, while only 6% would switch to a paid tier to avoid ads. OpenAI expects ads to help cover about $1.4tr in computing costs and projects that 8.5% of users (about 220 million people) will pay for an ad-free plan, while Anthropic is avoiding ads and targeting enterprise buyers. Trust in large tech companies is falling (from 32% to 24% since 2020), and concern is growing that ads could influence chatbot answers.
US ad spend is projected to increase 9.5% in 2026, with $9 billion in incremental dollars tied to midterms, the Winter Olympics, and the FIFA World Cup. Social and CTV will lead channel expansion at 14.6% and 13.8% YoY as buyers prioritize scalable, measurable environments. Economic risk persists with a one-in-three chance of recession, and 90% of buyers are concerned about tariffs, though those cutting spend due to tariffs fell from 45% to 30% YoY as budgets shift toward performance media. Rising acquisition costs have reduced its strategic weight by 10 points, yet 54% still rank new customer growth as a top objective.
Brands overcomplicate lead scoring by combining revenue potential and likelihood to close into one number. The better approach is to keep them separate. Use revenue potential to tier leads based on size or deal value, and use conversion likelihood driven by firmographic fit and engagement signals to score them. Focus on the few factors that actually correlate with closed deals.
The guidance represents a reversal of a decade of search best practices that rewarded tight segmentation and manual optimization. Google's position is that AI models can process real-time auction dynamics and contextual data at a scale no human team can match. Early adopters reportedly see improved conversion rates after consolidation, but marketers will give up granular reporting and tighter control for algorithmic performance gains that are harder to report and explain.
Sponsored
12% of people click through to verify AI recommendations. Evertune [10] shows you the publishers AI cites in your category, then helps you run programmatic ads on their sites. That means you can get seen in AI search even when your brand isn't mentioned in the answer. Try
FAQ sections help pages perform better for AI search because they pack high-similarity answers into very little text. Research that compared dense prose, FAQs, and structured content across five queries and multiple models showed that FAQ content had the highest similarity scores in almost all cases. AI systems often work under a grounding budget of about 2K words per query, so compact FAQ blocks are easier for them to process and reuse. Screaming Frog connects to the OpenAI API and can generate FAQs at scale by crawling pages and sending page text to a model, and one test produced 150%+ FAQs in about five minutes. AI-generated FAQs work best on longer pages and still need human review before publishing.
The Operators Growth Summit is a one-day free online event on February 18 that will feature 25+ DTC and ecommerce leaders sharing growth tactics. It includes three hands-on breakout sessions, plus a rapid-fire lightning panel of 10 experts. The event is designed for operators seeking real-world guidance, not theory.
AI coding tools now let marketers build free, lightweight satellite apps that replace static lead magnets and drive growth across the full funnel. These micro-apps give instant value, require little maintenance, and connect users directly to the main product through demos, calculators, benchmarks, and interactive reports instead of gated PDFs. Tools that once took weeks or months and full engineering teams can now be built in hours or days by one person. Companies report strong results such as high traffic, new signups, and cross-activation into core products from ungated tools like AI visibility and product benchmark apps. The approach works best when the core product relies on unique data, network effects, or trust, since simple point solutions can now be easily recreated.
McDonald's launched the limited-time McNugget Caviar kit before Valentine's Day after a viral high low trend. It was inspired by a $100 nugget and caviar dish at the 2025 US Open. The free kit included Baerii sturgeon caviar, crème fraîche, a mother-of-pearl spoon, and a $25 nugget voucher. The promotion hits a sweet spot by being free and culturally valuable. It lets a $5 meal chain wink at a $100 dish while inviting fans to play along.
Cloudflare's new feature serves AI agents your site content with fewer tokens, which saves resources and provides optimized content for both users and AI.
3 questions to ask during a discovery call include situational, gap, and future state.
Tips on LinkedIn features and analytics that increase reach and content discovery.
Sponsored
Everybody wants better alignment. In reality, this often devolves into tool sprawl, parallel work, and endless meetings where nothing gets done.
AI's variable response times fragment attention and weaken deep thinking. Leaders should address this as a coaching challenge and help teams manage AI wait time intentionally rather than relying on willpower.
Autonomous AI agents can now build features directly from structured intent, redefining how product teams work. PMs, engineers, designers, and QA move from manual execution to orchestrating intelligent systems.
Push-based planning fails in complex work because forecasts and estimates are unreliable. Pull-based systems work better by grounding plans in real capacity and actual progress rather than speculation.
Many agencies claim SaaS expertise but only design marketing sites or consumer apps, leading to costly mistakes in real B2B products. True SaaS specialization requires deep experience with multi-tenant systems, admin tools, and complex product workflows.
Sponsored
Unwrap's [10] customer intelligence platform brings all your customer feedback (surveys, reviews, support tickets, social comments, etc.) into a single view, then uses AI + NLP to surface the most actionable insights and deliver them straight to your inbox.
Even when product managers do everything right, fast and fragmented environments can overwhelm them. An environment strategy helps them decide where to focus and how deeply to engage so they can stay effective without burning out.
2026.02.13·amycmitchell.substack.com·5 min read·TLDR Product Wispr Flow's onboarding succeeds because it builds the right mental model, teaches new behaviors at the moment of action, prevents errors, and increases user investment through personalization. By making learning central to the experience, it turns onboarding into a powerful driver of adoption and habit formation.
PMs at The New York Times combine diverse backgrounds with strong strategy, leadership, and technical skills to advance trustworthy journalism. Their varied experiences help them serve readers, support journalists, and grow the business in a rapidly changing digital world.
AI coding tools like Claude Code let product managers turn ideas into working software and automate core workflows. For team collaboration and production deployment, platforms like Builder.io extend that capability into a shared, scalable system.
Top AI labs now expect PMs to run evals, code prototypes, understand model tradeoffs, and automate work with AI agents.
In high-stakes moments that determine a company's survival, leaders must stay calm, expand their options, trust their instincts, and commit fully once they decide.
Sponsored
Need a workflow builder you can trust with mission-critical tasks, but is also accessible to everyone on your team? Meet Tines, the intelligent workflow platform.
when you need more.
Elon Musk has restructured xAI following the exit of two of its co-founders earlier this week. xAI will now be organized into four core areas: Grok's chatbot and voice product, Coding, the Imagine video product, and Macrohard. The company plans to launch a new X Chat app and X Money in the coming months. It is currently looking to hire.
2026.02.12·links.tldrnewsletter.com·4 min read·TLDR Several highly anticipated Siri functions may be pushed back as testing has run into snags in recent weeks. Apple now plans to spread out new capabilities over several versions, so some features will be postponed until iOS 26.5 or iOS 27. One feature especially likely to slip is the expanded ability for Siri to tap into personal data. Another feature running behind schedule is voice-based control of in-app actions. While early support for these features exists, they don't function reliably in all cases.
2026.02.12·links.tldrnewsletter.com·9 min read·TLDR Tech is starting to turn to AI to probe the possibility of digitizing smell in the hopes of understanding and manipulating this sensory modality. Programming smell may illuminate many mysteries about the sense. It could also have many practical applications, for example, computational smell could help detect gas leaks, reduce reliance on intensive natural ingredients for perfume, or lead to the creation of entirely novel smells. This article takes a look at the state of the research and the future of the industry.
2026.02.12·press.asimov.com·37 min read·TLDR China's space program successfully carried out a flight of a new reusable booster and crew capsule late Tuesday. The Long March 10 rocket and Mengzhou spacecraft are core elements of China's lunar architecture. The US and China are racing to accomplish the next human landing on the Moon in a competition for national prestige and lunar resources. The successful test moves Chinese engineers closer to human-rating the Mengzhou crew capsule. A video recapping the test flight is available in the article.
2026.02.12·arstechnica.com·6 min read·TLDR Sponsored
2026.02.12·getunblocked.com·Sponsor·TLDR Most future distributed code will be AI-generated. This makes it important whether we ask AI to aim for ordinary imperative code or at frameworks where distributed contracts are part of the language surface. Targets like this help developers focus on the right questions rather than chasing shadows. AI can help build systems that fail less often if given a target that reflects distributed reality.
2026.02.12·jhellerstein.github.io·8 min read·TLDR Never let Claude write code until you have reviewed and approved a written plan. This separation of planning and execution prevents wasted effort, keeps users in control of architectural decisions, and produces significantly better results with minimal token usage. This post describes a workflow that follows this core principle. The workflow is radically different from how most people approach AI tools.
2026.02.12·boristane.com·15 min read·TLDR AI works, it improves predictably, and the richest institutions in history are committing trillions to it. It isn't a fad. The next few years will be disorienting in ways most people aren't prepared for. The people who will come out of this best are the ones who start engaging now.
2026.02.12·shumer.dev·28 min read·TLDR The story of the AI boom so far is one where the dominant narrative about AI is not what it has built, but the rate at which people are consuming it, the rate at which companies are spending on GPU farms, and the rate at which tools are being expensed. AI is everywhere in consumption and almost nowhere in output. We are spending unprecedented sums to invest in systems, and the primary product of that spending is the experience of spending it.
2026.02.12·links.tldrnewsletter.com·9 min read·TLDR