Featured Issues

A last look at Hemang Doshi's advice for AI, auditing, and privacyYou may have outsourced CIAM to the engineering team, but security still gets the call when there’s a breach. It’s time for you to take control, not the blame.Frontegg gives security teams direct control over the policies that safeguard your customer-facing application. No more waiting for developers to implement step-up MFA or manage compliance updates.Start Your Free TrialTake a look at the Security Suite directly#217: Privacy and YouAnother look at CISA and a survey of the landscapeWelcome to another_secpro!In cybersecurity, there's no such thing as standing still. While standing still might mean "going with the flow" in ordinary life, it means the very opposite when it comes to jousting with the adversary - indeed, standing still means "letting the flow go past you"! That's why we in the _secpro team are always pushing ourselves and pushing our readers to pick up ideas, develop skills, and stay above water in the rushing waves of "the flow"!That's why this week we are beginning a four-part series that looks into the deeds and needs of a CISA-trained professional - and, more importantly, how you can get to that plateau too. With the help of Hemang Doshi's fantastic book, we're taking the necessary steps to move from IT generalist or junior secpro into the higher echelons of auditing. Sound good? Check out this week's excerpt: Data Privacy Program and Principles.Check out _secpro premiumIf you want more, you know what you need to do: sign up to the premium and get access to everything we have on offer. Click the link above to visit our Substack and sign up there!Cheers!Austin MillerEditor-in-ChiefAdvance your technical career with actionable, practical solutionsAWS re:Invent 2025 Las VegasTransform your skills at AWS re:Invent 2025. Master new AWS services, join immersive workshops, and network with top cloud innovators at AWS re:Invent 2025. As a re:Invent attendee,you'll receive 50% discount code towards any AWS Certification exam.Our 2025 event catalog is now available!Explore the EventHere's a little meme to keep you going...Source: RedditThis week's articleData Privacy Program and PrinciplesAI is revolutionizing various industries, including auditing. Traditionally, auditing has been a manual and time-consuming process, requiring auditors to sift through large volumes of data to identify discrepancies and ensure compliance. However, with the advent of AI, the audit process is becoming more efficient, accurate, and insightful. AI can analyze vast amounts of data quickly, identify patterns, and even predict potential risks, making it an invaluable tool in modern auditing.Read the rest here!News BytesCisco ASA / FTD Zero-Days Under Active Exploitation: On 25 September, Cisco and CISA published security advisories confirming that multiple zero-day vulnerabilities affecting Cisco ASA / FTD (Firewall, VPN) products are being actively exploited. Two of these (CVE-2025-20333, CVE-2025-20362) were confirmed to have been exploited in the wild.Threat actors have leveraged advanced evasion techniques (disabling logs, intercepting CLI commands, modifying boot processes) and deployed bootkits such as RayInitiator combined with malware (e.g., LINE VIPER) to persist across reboots and firmware upgrades. The urgency prompted CISA to issue an Emergency Directive 25-03, mandating U.S. federal agencies to inventory, assess, and mitigate vulnerable Cisco devices.Continued Attack Campaign on Cisco Firewalls (Rommon / Bootkit-level Persistence) (PDF): Following the zero-day disclosures, deeper forensics revealed that the adversaries are not merely exploiting web/VPN logic flaws, but targeting the ROM Monitor (ROMMON) / boot environment of ASA devices. The RayInitiator bootkit persists in the boot chain, and it loads LINE VIPER, a malware module that can intercept commands, bypass VPN AAA, suppress logs, and embed itself into core ASA processes (e.g. lina). Some devices lack Secure Boot / Trust Anchor support, making them more vulnerable. These mechanisms impede forensic detection and complicate patching strategies — for example, even after reboots or upgrades, malicious modules can survive.Scattered Spider: Retail Service Desk Exploits Renewed Focus: Throughout the week, multiple analyses surfaced reaffirming that the hacking collective Scattered Spider (aka UNC3944 / Octo Tempest) is continuing to rely heavily on social engineering of service desks / help desks to gain initial footholds in enterprise networks. A new PDF—Cross-Sector Mitigations: Scattered Spider—jointly produced by sector cyber-information shares, outlines updated TTPs (tactics, techniques, procedures) and countermeasures for financial services, IT/retail, health, etc. In one prominent case, attackers impersonated internal staff, tricked the helpdesk into resetting MFA / disabling controls, and escalated privileges inside M&S / Co-op systems. Forensic Visualization Toolkit: Enhancing Threat Hunting: In a freshly published academic work (11 September 2025), researchers present “Enhancing Cyber Threat Hunting – A Visual Approach with the Forensic Visualization Toolkit”. The toolkit offers interactive visualizations of forensic and telemetry data (network, file access, process graphs) to assist threat hunters in spotting anomalies that may evade automated detection systems. The authors argue that combining human analytical insight with visualization accelerates detection of stealthy threats, especially those embedded in normal-looking activity windows.The paper includes realistic case studies and performance comparisons, making it a timely reference for SOC / IR teams aiming to ramp threat‐hunting maturity.Burnout in Cybersecurity: A Strategic Risk Report: While not a direct breach event, a notable paper published earlier in 2025 — “A Roadmap to Address Burnout in the Cybersecurity Profession” — has gained renewed attention this week in security circles. The work synthesizes findings from a multi-disciplinary workshop involving practitioners, academics, and ex-NSA cyber operators. It outlines the human, organizational, and workflow stresses contributing to attrition and mental fatigue, and presents a roadmap of interventions (training, rotation, psychological support, team-based structures) to mitigate erosion of security capacity. Given current pressure on SOC/IR teams (e.g. responding to high-tempo incidents like the Cisco zero-days), this issue is increasingly treated as a strategic risk in cybersecurity planning.Digital Forensics & Risk Mitigation Strategy for Modern Enterprises: Another academic contribution gaining traction is “Comprehensive Digital Forensics and Risk Mitigation Strategy for Modern Enterprises”, published February 2025. The paper walks through a simulated case of a large identity/data-analytics firm under attack and develops an integrated strategy covering pre-incident readiness (forensic architecture design, monitoring), live response, post-incident lessons, and regulatory compliance.It emphasizes adaptive AI/ML techniques, integration of threat intelligence into forensics workflows, and continuous “forensic readiness” as a discipline. In the context of emerging threats (e.g. boot-level persistence, identity-based service desk attacks), the paper serves as a robust blueprint for mature enterprise response programs.This week's academiaAdversarial Machine Learning: A Taxonomy and Terminology: A comprehensive NIST report that builds a clear taxonomy and standardized terminology for adversarial machine learning (AML). It describes attacker goals and capabilities across ML life-cycles, categorizes AML attack and defense types, and outlines current technical and measurement challenges for trustworthy AI in security-sensitive systems. Highly cited and used as a baseline by both researchers and practitioners.(A. Vassilev et al. NIST Trustworthy & Responsible AI group).On Adversarial Attack Detection in the Artificial Intelligence Era: Survey/analysis of detection techniques for adversarial attacks on ML models, contrasting classic concealment/malware tactics with modern adversarial-example threats. The paper evaluates state-of-the-art detection approaches and points to gaps where attackers are leveraging large models and automation to evade defenses. Useful for defenders designing layered ML security. (N. Al Roken and collaborators).A Defense-Oriented Model for Software Supply Chain Security: Introduces the AStRA graph-based model (Artifacts, Steps, Resources, Principals) to represent software supply chains and reason about security objectives and defenses bottom-up. Applies the model to case studies and maps past supply-chain attacks to show where defenses succeed or fail — a practical roadmap for research and industry focusing on supply-chain mitigations (SBOMs, build integrity, provenance, etc.). (E. A. Ishgair and coauthors).Securing Automotive Software Supply Chains: NDSS paper that examines unique risks in automotive software supply chains (ECUs, OTA updates, third-party components). It evaluates real automotive update pipelines, shows practical attack scenarios, and recommends defenses tailored to the automotive context (signing, reproducible builds, hardened update channels). Very relevant given recent high-profile industrial supply-chain incidents. (Marina Moore, Aditya Sirish A. Yelgundhalli, Justin Cappos).Managing Deepfakes with Artificial Intelligence: Introducing a Business/Privacy Calculus: Academic analysis of deepfake threats and defenses from both technical and socio-economic angles. Proposes an AI-assisted detection/mitigation framework and a privacy/business calculus for organizations to evaluate risks vs. countermeasure costs (useful for enterprises facing deepfake-enabled fraud or reputational attacks). Timely as synthetic media use explodes. (G. Vecchietti and collaborators).*{box-sizing:border-box}body{margin:0;padding:0}a[x-apple-data-detectors]{color:inherit!important;text-decoration:inherit!important}#MessageViewBody a{color:inherit;text-decoration:none}p{line-height:inherit}.desktop_hide,.desktop_hide table{mso-hide:all;display:none;max-height:0;overflow:hidden}.image_block img+div{display:none}sub,sup{font-size:75%;line-height:0} @media (max-width: 100%;display:block}.mobile_hide{min-height:0;max-height:0;max-width: 100%;overflow:hidden;font-size:0}.desktop_hide,.desktop_hide table{display:table!important;max-height:none!important}}

Meta’s new specs, OpenAI’s big spend, and other AI adventures AI_Distilled #115: What’s New in AI This Week Hello there! Welcome to your weekly roundup of all things newsworthy in tech. Grab your coffee, settle in, and let’s dive into the highlights from the past week. LLM Expert Insights, Packt 📈LATEST DEVELOPMENT Google puts Chrome on an AI caffeine rush Google’s Chrome browser got 10 new AI-powered features led by Google’s Gemini model. It can now summarize webpages, explain multiple tabs, and even autonomously book appointments or grocery orders (no kidding). The browser also gains a freakishly good memory—just ask where you saw that walnut desk last week, and Chrome will pull up the page. Efficiency level: Max. ChatGPT falls for a Gmail trick, OpenAI marks it as resolved Even AI isn’t safe from sneaky hacks. Radware revealed a crafty prompt injection attack that tricked OpenAI’s Deep Research agent into exfiltrating Gmail data. In plain speak: a bad email with hidden instructions made the AI steal email secrets. OpenAI patched the hole with more safety checks, but it’s a reminder that letting an AI rummage through your inbox can be a minefield of surprises (ouch). Although the issue was reported in June, OpenAI has reportedly marked it as resolved. Advance your technical career with actionable, practical solutions | AWS re:Invent 2025 Las Vegas Transform your skills at AWS re:Invent 2025. Master new AWS services, join immersive workshops, and network with top cloud innovators at AWS re:Invent 2025. As a re:Invent attendee,you'll receive 50% discount code towards any AWS Certification exam.Our 2025 event catalog is now available! EXPLORE THE EVENT Luma AI’s Ray3: Lights, Camera, AI! Startup Luma AI unveiled Ray3, an AI toolkit that brings Hollywood-level wizardry to your phone. Integrated with Adobe, Ray3 can generate HDR video (10-, 12-, 16-bit color) and even turn boring SDR footage into vivid HDR. Its built-in reasoning engine lets creators sketch out camera movements or scene edits, and the AI dutifully follows multi-step instructions. It’s like having a tiny James Cameron in your pocket, minus the ego. Meta’s smart glasses Meta’s Connect 2025 event, Zuck & Co. pivoted from metaverse musings to real hardware. While the glasses encountered problems during the live demo due to a race condition bug, the future is not as bleak. The Ray-Ban Meta smart glasses priced at $799, rocking a microLED display that projects messages, maps, and more right in your field of view. You control these AR specs with a neural wristband (bye-bye, clunky controllers). It’s the closest thing to wearing Tony Stark’s tech. OpenAI’s Manhattan project for AI Sam Altman is going big. OpenAI is teaming up with Oracle, SoftBank, and Nvidia to build out an AI super-infrastructure that makes current data centers look like Lego blocks. They’re planning five new U.S. data centers (bringing as much power as 7 nuclear reactors worth of energy!) and exploring a bold new “GPU leasing” deal worth $100B with Nvidia. In short, OpenAI wants endless computing power on tap, betting that in the AI race, bigger is better (and necessary). Oracle bets billions on cloud AI Larry Ellison must be feeling lucky. Why? Well, it is reported that Oracle is reportedly close to a $20 billion deal with Meta to host and train Meta’s AI models. This comes right after Oracle’s whopping $300B contract with OpenAI and a new partnership with Elon Musk’s xAI. The strategy? Offer faster, cheaper cloud infrastructure to undercut Amazon and Microsoft. If this pays off, Oracle’s cloud might go from underdog to top dog in the AI era. Bold move, Larry. Musk’s xAI drops a game-changer Elon Musk’s new AI venture, xAI, just launched a model called Grok 4 Fast that claims GPT-5 level smarts at a fraction of the cost. We’re talking near top-tier reasoning benchmarks with 98% lower token costs. It achieves this by cutting out “thinking overhead” and streamlining how it chews through data. Translation: powerful AI answers, cheap enough to deploy en masse. It’s Musk’s way of saying “Competition, bring it on.” Brain implants: Neuralink’s next step As per a Bloomberg report, Elon’s neuro-lab: Neuralink is gearing up for its first human trials this October after getting FDA’s nod. The company’s implantable chip can translate thoughts to text, initially aimed to help paralyzed patients communicate. Long term, Musk envisions people using thoughts to control computers and even converse with AI—because typing is so 2020, right? It’s equal parts exciting and sci-fi-level eerie. Alibaba’s model mega-mix Not to be outdone, Alibaba unveiled its Qwen3 AI stack with a twist: Mixture-of-Experts (MoE) models at trillion-parameter scale. The system can tap into 512 expert models but activates just a handful per query for super efficiency. End result? Over 10× throughput improvement and support for ridiculously long context (think entire novels in one prompt). Two 80B-version models lead the charge—one tuned for chatty assistants, another for complex reasoning. In the AI model arms race, Alibaba just loudly entered the chat. Microsoft’s developer boost (and cool chips) Redmond had a productive week too. Microsoft is hunting down pesky legacy code with new Copilot-powered agents that not only find problems in old .NET/Java code, but also auto-generate fixes, unit tests, and containerize apps. Early trials showed dramatic wins – an Xbox team cut migration effort by 88% and Ford saw a 70% reduction in update time. On another front, Windows 11 now comes with built-in support for running AI models (ONNX runtime) across CPUs, GPUs, and specialized NPUs from various vendors. And about “cooling chips from the inside out”? Microsoft researchers are exploring liquid cooling inside chips to solve overheating as AI silicon gets hotter (literally). The future: faster chips that keep their chill. START YOUR FREE TRIAL EXPERT INSIGHTS Introduction to chunking with GPT-4o In generative AI workflows, the way data is prepared has a direct impact on model effectiveness. Rather than relying solely on rule-based chunking methods, this tutorial introduces an approach where GPT-4o itself is used to intelligently divide unstructured content into meaningful segments. This strategy supports a Retrieval-Augmented Generation system and enables it to more effectively retrieve relevant context. Why chunking matters in GenAI systems Traditional chunking methods often split documents based on arbitrary rules, such as paragraph breaks or token counts, which may cut through semantically meaningful units. In contrast, intelligent chunking enables each piece of data to carry a coherent message. This is particularly important when chunks are embedded into a vector database like Pinecone for retrieval. If a query surfaces a partial or poorly segmented chunk, the generated response may lack clarity or precision. Using GPT-4o for semantic chunking GPT-4o is employed not just for generation but also as a semantic analyzer. The model receives the full unstructured text, such as a company memo or technical note, and is prompted to divide it into logically structured chunks, each roughly 50–100 words in length. This is achieved by setting up a system message instructing the model to act as a chunking assistant, followed by a user message containing the text to split. Consider this system message: "You are an assistant skilled at splitting long texts into meaningful, semantically coherent chunks of 50–100 words each. Split the following text into meaningful chunks..." Once the prompt is issued, GPT-4o returns a response with double newlines separating each chunk. The program parses the response by splitting on these newline markers. The result is a list of discrete, meaningful units that are ready to be embedded. This workflow is especially useful for processing internal company data, like executive summaries or operational notes, where nuance matters. This method shines when the data includes complex thoughts, mixed formats, or narrative elements. For simpler documents like lists or spreadsheets, rule-based chunking might be more efficient. However, for nuanced tasks where meaning spans sentences or paragraphs, GPT-4o’s semantic awareness offers a significant advantage. By integrating GPT-4o into the chunking process, generative AI systems can store and retrieve content in a more meaningful way. Each chunk becomes a high-value data unit, tailored for precision recall within an RAG pipeline. This intelligent preprocessing step reinforces the larger GenAISys vision, building systems that retrieve not just data, but context-rich, purpose-aligned information. Built something cool? Tell us. Whether it's a scrappy prototype or a production-grade agent, we want to hear how you're putting generative AI to work. Drop us your story at nimishad@packtpub.com or reply to this email, and you could get featured in an upcoming issue of AI_Distilled. 📢 If your company is interested in reaching an audience of developers and, technical professionals, and decision makers, you may want toadvertise with us. If you have any comments or feedback, just reply back to this email. Thanks for reading and have a great day! That’s a wrap for this week’s edition of AI_Distilled 🧠⚙️ We would love to know what you thought—your feedback helps us keep leveling up. 👉 Drop your rating here Thanks for reading, The AI_Distilled Team (Curated by humans. Powered by curiosity.) *{box-sizing:border-box}body{margin:0;padding:0}a[x-apple-data-detectors]{color:inherit!important;text-decoration:inherit!important}#MessageViewBody a{color:inherit;text-decoration:none}p{line-height:inherit}.desktop_hide,.desktop_hide table{mso-hide:all;display:none;max-height:0;overflow:hidden}.image_block img+div{display:none}sub,sup{font-size:75%;line-height:0} @media (max-width: 100%;display:block}.mobile_hide{min-height:0;max-height:0;max-width: 100%;overflow:hidden;font-size:0}.desktop_hide,.desktop_hide table{display:table!important;max-height:none!important}}