Bytedance's Open-Weight Helios Model Brings Minute-Long AI Video Generation Close to Real Time
Bytedance released Helios, a 14‑billion‑parameter video model that generates minute‑long clips at 19.53 frames per second on a single H100 GPU. The distilled version slashes inference steps from 50 to three, achieving speeds comparable to much smaller 1.3 B models while maintaining superior visual quality. Helios uses a unified architecture that supports text‑to‑video, image‑to‑video, and video‑to‑video generation, automatically switching based on context. The model and weights are openly available on GitHub and Hugging Face for research use.
AI Can Mass-Unmask Pseudonymous Accounts, Research Paper Finds
Researchers from ETH Zurich and Anthropic demonstrated that large language models can deanonymize pseudonymous users on platforms such as Hacker News and Reddit. In controlled experiments the AI agent correctly linked two‑thirds of anonymous profiles to real identities, a task...
Anthropic Turns Claude Code Into a Background Worker with Local Scheduled Tasks
Anthropic has expanded its Claude Code AI coding assistant with a new "/loop" command that lets users schedule recurring background tasks. The feature supports standard cron expressions, allowing intervals from minutes to days, and can handle up to 50 tasks...
Anthropic's New Marketplace Lets Enterprise Customers Spend Their Existing AI Budget on Third-Party Tools
Anthropic announced the Anthropic Marketplace, a storefront where enterprise customers can purchase third‑party applications built on Anthropic’s AI models. Launch partners include Snowflake, Harvey, and Replit. The company will not charge commissions and allows customers to allocate part of their...
Karpathy’s March of Nines Shows Why 90% AI Reliability Isn’t Even Close to Enough
Andrej Karpathy’s “March of Nines” highlights that achieving 90% AI reliability is only the first step; each additional nine of uptime demands comparable engineering effort. In multi‑step agentic workflows, the probability of success compounds exponentially, turning a seemingly robust demo...
Google Launches TensorFlow 2.21 And LiteRT: Faster GPU Performance, New NPU Acceleration, And Seamless PyTorch Edge Deployment Upgrades
Google released TensorFlow 2.21, promoting LiteRT from preview to a production‑ready on‑device inference framework that replaces TensorFlow Lite. LiteRT delivers a 1.4× speed boost on GPUs and introduces unified NPU acceleration for edge hardware. The update expands low‑precision operator support, adding...
ZyG
ZyG has launched an AI‑driven Growth Engine designed to help early‑stage direct‑to‑consumer (DTC) startups scale their eCommerce operations. The platform delivers product‑market fit analysis, centralized access to third‑party tools, and automated workflows that streamline growth initiatives. ZyG’s leadership team combines...
Anthropic Launches Claude Marketplace, Giving Enterprises Access to Claude-Powered Tools From Replit, GitLab, Harvey and More
Anthropic unveiled Claude Marketplace, a limited‑preview platform that lets enterprises apply existing Claude spend toward third‑party tools built on its models. The offering aggregates partners such as GitLab, Replit, Snowflake, Harvey and Rogo, and consolidates invoicing to streamline procurement. By...
State CIOs Have a New Top Priority in 2026
State chief information officers have elevated artificial intelligence to the top of their 2026 agenda, displacing cybersecurity after a 12‑year run. The NASCIO survey shows more than 90% of states are piloting generative AI projects and have established task forces...
US AI Boom Creates Big Market Opportunity for Renewable Gas
The United States is experiencing a rapid surge in artificial‑intelligence workloads, prompting a projected 30 percent increase in electricity demand by 2030. To satisfy this growth, utilities are turning to renewable natural gas (RNG) as a low‑carbon, dispatchable fuel for data...
MCP C# SDK 1.0 Arrives with Improved Authorization Server Discovery
Microsoft released the 1.0 version of its official C# SDK for the Model Context Protocol (MCP). The SDK implements the latest 2025‑11‑25 MCP specification, adding enhanced authorization server discovery, icon metadata, and client ID metadata documents. It also supports tool...
Highmark Health Generates $28M in Value with Google AI
Highmark Health, a Pittsburgh‑based payer‑provider, reported $27.9 million in generated value for 2025 from its Google Cloud‑built AI assistant, Sidekick. Employees have prompted the tool more than six million times, expanding active use cases to 74, up from 31 the previous...
Can AI-Driven Security Solutions Fit Small Business Budgets
Non‑Human Identities (NHIs) are machine credentials that secure cloud‑based interactions across sectors such as finance, healthcare, and travel. Managing the full NHI lifecycle—discovery, classification, monitoring, threat detection, and remediation—provides a holistic security posture far beyond point solutions. AI‑driven analytics now...
Significant Department Of Air Force Readiness Boost May Come From Process Automation, Epsilon3 Exec Says
Epsilon3, a Los Angeles‑area startup, has secured contracts with the U.S. Air Force and Space Force to automate manual testing and launch procedures using AI‑driven software. The platform digitizes test plans, tracks aircraft and satellite testing, and provides real‑time analytics to...
Staying Ahead of AI-Driven Cyber Threats
At the Zscaler Public Sector Summit, Idaho National Laboratory’s CISO Robert Roser warned that artificial intelligence is accelerating the sophistication of phishing and ransomware attacks targeting government and critical‑infrastructure networks. He noted AI lowers the technical barrier for less‑skilled hackers,...
OpenAI Quietly Scales Back Direct Checkout Plans for ChatGPT
OpenAI has quietly scaled back its direct checkout feature for ChatGPT, postponing a broad consumer rollout and limiting the capability to a small group of pilot merchants. The company cited technical integration hurdles and regulatory scrutiny as primary reasons for...
New KV Cache Compaction Technique Cuts LLM Memory 50x without Accuracy Loss
MIT researchers introduced Attention Matching, a fast KV‑cache compression technique that can shrink large language model memory footprints by up to 50 times without measurable accuracy loss. The method preserves attention output and attention mass by fitting compressed keys and values...
OpenAI Introduces Codex Security in Research Preview for Context-Aware Vulnerability Detection, Validation, and Patch Generation Across Codebases
OpenAI has rolled out Codex Security, an application security agent, in research preview for ChatGPT Enterprise, Business, and Edu customers via Codex web. The tool builds a project‑specific threat model, validates vulnerabilities in sandboxed environments, and generates context‑aware patches. In...
AI’s Impact on Warfighting Starts With Logistics
Defense Logistics Agency CIO Adarryl Roberts outlined a "connected enterprise" framework that embeds AI across supply‑chain functions, from procurement to finance. The agency is constructing a data‑mesh backbone to feed "control tower" dashboards, giving senior leaders real‑time logistics visibility. Deterministic...
Tencent Games Unveils AI Tools and Developer Sessions for GDC 2026
Tencent Games announced a robust GDC 2026 program featuring more than 20 sessions and a technology showcase of its latest AI‑powered development tools. The company will exhibit six internal solutions—including VISVISE, MagicDawn, ACE anti‑cheat, WeTest, ASI World, and AI‑generated Kung Fu...
Google PM Open-Sources Always On Memory Agent, Ditching Vector Databases for LLM-Driven Persistent Memory
Google has open‑sourced an Always On Memory Agent that eliminates traditional vector‑database retrieval in favor of a large‑language‑model‑driven memory layer. Built on the Agent Development Kit and powered by the low‑cost Gemini 3.1 Flash‑Lite model, the agent continuously ingests, consolidates, and serves...
Why Replacing Anthropic at the Pentagon Could Take Months
The Department of Defense has given Anthropic six months to remove its Claude model from classified networks, citing it as a supply‑chain risk. While swapping the model technically takes minutes, retraining personnel and re‑engineering workflows will take months. The move...
DSG Expands AI and Ecommerce Across Its Distribution Brands
Distribution Solutions Group (DSG) is embedding ecommerce and AI across its three main brands—Lawson Products, Gexpro Services and TestEquity—to boost sales efficiency and capture new customers. In Q4, Lawson’s online channel posted an 18% revenue increase, with over 30% of...
I Turned My Linux Terminal Into a Local AI Assistant and It’s so Useful
The author demonstrates how to turn a Linux terminal into a local AI assistant using Ollama and the 2 GB Llama 3.2 model, running entirely on a CPU‑only laptop. By adding two Bash functions—`ask` and `explain`—the model can interpret commands, man pages,...
Virginia Dignum on The AI Paradox
Virginia Dignum’s new book *The AI Paradox* argues that the growing capabilities of artificial intelligence actually highlight the irreplaceable value of human creativity, moral judgment, and responsibility. She frames AI’s biggest challenges as enduring paradoxes—tensions between efficiency and control, innovation...
Oracle Health Embedding AI to Improve Care and Increase Efficiency
Oracle Health will showcase a suite of AI‑driven solutions at HIMSS26, including its Clinical AI Agent that has been adopted by more than 300 organizations and has saved doctors over 200,000 hours of documentation time. The company is also unveiling...
Q&A: Sanford Health Bets on AI, Virtual Care to Expand Rural Healthcare Access
Sanford Health, the nation’s largest rural provider serving over two million patients across seven states, leveraged a $350 million donation to accelerate its virtual‑care program. The organization opened a Sioux Falls virtual‑care center that educates clinicians on digital bedside manner and hosts an...
Longevity and Disease Insights Now in 20/20 BioLabs Blood Test
20/20 BioLabs has launched OneTest for Longevity, a lab‑developed blood test that combines inflammatory biomarkers, lifestyle data, and AI to deliver personalized aging and chronic disease risk insights. The platform leverages IBM's watsonx.ai and the University of South Carolina's Dietary...
Robotics Ethicist Calls for Stronger US Guardrails as Automation Accelerates
Robotics ethicist Kate Darling warned that U.S. policy is falling behind the rapid deployment of automation in warehouses and factories. She argued that profit-driven decisions, not technical capability, will shape the future of work unless stronger regulatory guardrails are introduced....
Podscribe Launches AI Chatbot and Introduces YouTube SmartModeling to Improve Podcast & Video Ad Measurement
Podscribe unveiled two major product upgrades: an AI‑powered chatbot that lets advertisers query the platform’s podcast data in natural language, and YouTube SmartModeling, a new attribution framework for simulcast ads. SmartModeling replaces the old practice of mirroring podcast conversion rates...
Channel Brief: Automation, AI, and Compliance Expand MSP Service Opportunities
Managed service providers are rapidly adopting automation, AI‑enabled platforms, and compliance solutions to reduce manual effort and meet rising regulatory demands. Vendors such as GitLab, LogicMonitor, ConnectSecure, Cork Cyber, and Monjur are launching AI‑driven tools that let MSPs package DevSecOps...
DataDome, Botify Partner on Agentic Commerce Control
DataDome and Botify announced a partnership to help retailers manage the emerging agentic commerce ecosystem, combining bot‑trust management with AI‑search optimization. Joint research shows 73% of consumers have used AI assistants and AI bot traffic surged 5.4‑fold in 2025, while...
Arbital Health Sees Rapid Adoption of Actuarial AI
Arbital Health announced rapid market adoption of its Merlin AI actuarial assistant, launched in October 2025. Leading payers and providers such as Arkos Health and CommuniCare have integrated the platform to monitor value‑based care contracts, forecast financial impact, and identify...
The Decline of the Cover Letter in the AI Era
Professor Judd Kessler argues that artificial intelligence is diminishing the signaling value of traditional cover letters. AI-driven resume parsers focus on quantifiable data, making narrative essays less relevant. As a result, recommendations, personal networks, and demonstrable achievements are becoming the...
AI Agents Pose Untold Risk to Humanity. We Must Act to Prevent that Future | David Krueger
David Krueger warns that AI agents are evolving from assistants to autonomous actors, citing Moltbook—a platform where AIs converse, form a religion, and even propose humanity's purge. Real‑world mishaps, such as an OpenClaw agent deleting a researcher’s inbox, illustrate how...
North Korean Agents Using AI to Trick Western Firms Into Hiring Them, Microsoft Says
Microsoft’s threat‑intelligence unit warned that North Korean state‑backed actors are leveraging AI tools—voice‑changing software, face‑swap apps, and synthetic name generators—to fabricate credible identities and secure remote IT positions with Western firms. Once hired, the impostors funnel salaries back to the...
CallRail Integrates with Yardi
CallRail has linked its AI‑driven Voice Assist platform with Yardi’s property‑management suite, enabling real‑time data retrieval during leasing calls. The integration lets the voice agent answer availability and pricing questions, qualify prospects, and push call insights back into Yardi’s system....
The Economics of Enterprise AI: What the Forrester TEI Study Reveals About Microsoft Foundry
Forrester’s Total Economic Impact study shows Microsoft Foundry delivers a 327% return on investment over three years for a composite enterprise, driven largely by developer productivity gains. The platform boosts technical team efficiency by up to 35%, cuts infrastructure waste,...
AI Offers More Intelligent Building Security Alerts, Johnson Controls Expert Says
Johnson Controls' vice president of innovation Greg Parker says AI can make building security alerts faster, more accurate, and actionable, detecting camera misalignments, crowd formation, and anomalies in milliseconds. The Connected Security platform now delivers probabilistic alarm insights and automated...
Cardiology Now Has More than 200 FDA-Cleared AI Algorithms
An updated FDA registry shows cardiology now hosts more than 200 cleared AI algorithms, including 140 directly listed under the specialty and an additional 63 imaging‑focused tools. The total number of FDA‑cleared clinical AI applications reached 1,451, with radiology still...
How AI and Data Drive Compliance, Consistency and Scale
Europe’s e‑commerce market remains fragmented by languages, tax regimes and platform‑specific rules, forcing brands to adopt sophisticated back‑office operations. Luke Elliott of Pattern highlights that AI and real‑time data are now essential for meeting compliance, maintaining consistency and scaling across...
FDA Clears AI-Enabled Cardiac MR Planning Technology From Philips
Philips received FDA clearance for SmartHeart, an AI‑driven planning solution that automates cardiac magnetic resonance (CMR) exam setup in under 30 seconds. The system configures 14 standard views, reduces patient breath‑holds by up to 75%, and embeds motion‑correction (Cardiac MoCo)...
MinIO Integrates Delta Sharing Open Protocol for Seamless Access to Enterprise Data
MinIO has launched AIStor Table Sharing, embedding the Delta Sharing open protocol directly into its AIStor object store. The feature lets enterprises expose on‑premises data to Databricks in real time, eliminating the need for costly data replication. Built on Iceberg...
Kroger Names 84.51° Chief to Oversee AI Efforts
Kroger has created a new chief data and AI officer role, appointing Milen Mahadevan, president of its 84.51° analytics unit, to lead the effort. Mahadevan will consolidate all data and AI teams under his oversight while continuing to run 84.51°....
Liz Reid on Where Google Search Ends and Gemini Begins
Google Search head Liz Reid discussed AI’s impact on the flagship product. She noted that AI results have cut traffic to major tech sites, but framed the change as “expansionary.” When asked about merging Search with Gemini, she said she...
AI-Enabled MRI Scanner Gives South Shore Health New Approach to Imaging
South Shore Health has installed an AI‑enabled MRI scanner that accelerates scan times and delivers higher‑resolution images for physicians. The system leverages machine‑learning reconstruction to cut patient throughput by up to 30 percent while enhancing diagnostic confidence. CIO Dr. Sam...
How AI Created the ‘Entry-Level Squeeze’ and 3 Solutions to This Talent Crisis
AI’s rapid adoption is eliminating routine tasks, prompting many firms to slash early‑career roles. Avature’s AI Impact Report finds 76% of HR leaders anticipate a sharp drop in entry‑level hiring, creating a structural "entry‑level squeeze." The trend threatens the apprenticeship...
Optum Real, Microsoft Partner on AI for Claims and Reimbursement
Optum Real has teamed with Microsoft to accelerate claims and reimbursement using Azure, Dragon Copilot and Microsoft Foundry. The joint solution embeds AI‑driven coverage predictions, documentation assistance and prior‑authorization support into Optum Real’s real‑time platform. Pilot data show up to...
Building the Foundations of AI Cities: Lessons in Infrastructure, Energy, and Efficiency
AI’s exponential compute needs are outpacing traditional data‑center infrastructure, especially in legacy hubs like Northern Virginia. Operators are now targeting locations where clean power, talent, and renewable resources converge, birthing “AI cities” that embed digital infrastructure into urban planning. Rio...
AI Exploits, Cloud Breaches, and Identity Gaps Define This Week’s Cybersecurity Landscape
This week’s cybersecurity briefing highlighted a surge of AI‑driven exploits, including the unpatched MS‑Agent flaw in ModelScope and a patched prompt‑injection bug in Perplexity’s Comet browser. Critical infrastructure suffered high‑impact vulnerabilities such as Juniper PTX routers allowing unauthenticated root takeover...