Today's AI Pulse
Anthropic shuts down Claude Fable 5 and Mythos 5 after US export‑control warning
Anthropic disabled access to its Claude Fable 5 and Claude Mythos 5 models after the Trump administration, via the Commerce Department, warned that foreign nationals could exploit the systems. The move followed alerts from Amazon and other firms about a potential jailbreak that could aid cyber‑security attacks. Anthropic declined the administration’s request for a voluntary shutdown but complied with the order.
Also developing:
By the numbers: AMD acquires AI memory‑optimization pioneer Mext
Diligent AI Raises $2.5M to Support KYC and AML Teams with AI Agents
London‑based Diligent AI, a Y Combinator‑backed startup, has closed a $2.5 million seed round led by Speedinvest with participation from fintech investors and strategic angels. The company builds autonomous AI agents that automate repetitive KYC and AML tasks such as risk profiling, adverse‑media analysis, and sanctions screening. Its platform is already deployed in Europe, the United States, the Middle East and Japan, delivering measurable efficiency gains and more consistent decision‑making. The new capital will fund engineering expansion and accelerate rollout across the UK and broader European market.
AI Revisited - Part 2
In this episode of Rework, host Kimberly Rhodes and 37signals CEO Jason Fried discuss how AI tools like Claude and ChatGPT have become integral to their daily workflow. Jason shares concrete examples of using AI as a writing assistant to...
AI-Music Artists Get Their Own Accelerator From Prose Ventures
Prose Ventures has launched the AI Music Creators Accelerator, an eight‑week program that supports human musicians using generative AI tools. The inaugural cohort presented a diverse demo day on February 27, streamed publicly on Twitch and later on YouTube. Founder...
ArmorCode AI Exposure Management Identifies, Governs, and Reduces Shadow AI Risk
ArmorCode introduced AI Exposure Management (AIEM) on its Agentic AI Platform, expanding its unified exposure management suite. AIEM continuously ingests AI usage signals from security tools, creating a centralized inventory and assigning ownership to mitigate shadow AI. The solution offers...
Introduce an AI Policy Before It’s Too Late
Accounting firms are urged to formalise AI usage policies as artificial intelligence becomes integral to audit, tax and advisory work. The Finance, Accounting & Bookkeeping Show in Birmingham will host a dedicated programme on AI threats and opportunities, highlighting the...
Elon Musk Is Stunned by Alibaba’s New Qwen 3.5: Why the 9B Model Is Outperforming AI Giants 10x Its Size
Alibaba unveiled the Qwen 3.5 series, offering open‑source models ranging from 0.8 B to 9 B parameters with multimodal text‑and‑image capabilities. The flagship Qwen 3.5‑9B claims benchmark performance comparable to much larger models such as GPT‑oSS‑120B, ChatGPT and Gemini, especially in logical reasoning and...
Arkose Device ID Uses AI to Recognize Devices Across Changing Fingerprints
Arkose Labs unveiled the latest version of Arkose Device ID within its Arkose Titan platform, adding AI‑driven similarity analysis to traditional exact‑match identification. The enhancement allows persistent device recognition even as fingerprints evolve, reducing fraud from identity‑fragmentation attacks while keeping...
Extra #3 - The Prompt Injection Defense Playbook
The post outlines a premium playbook for defending Large Language Models against prompt injection, a semantic attack that tricks AI into violating its own constraints. It categorizes three primary attack vectors—role‑playing jailbreaks, hidden‑text payloads, and direct overrides—and proposes a multi‑layered...
News in Brief: Robotic Surgery Milestone at North Bristol, AI and RPA at Midlands Partnership, AI-Assisted Echocardiography in Cheshire and...
North Bristol NHS Trust performed its 10,000th robotic-assisted surgery, underscoring a decade of minimally invasive growth across multiple specialties. Meanwhile, Alder Hey Children’s NHS Foundation Trust expanded its 360° virtual tour to a full‑hospital map covering 313,585 square feet, helping...
Customer Joint Value Creation with Bimbo Bakeries USA's Jeff Hendrix
In this episode, Jeff Hendricks, Chief Customer Officer of Bimbo Bakeries USA, discusses how the company is leveraging real‑time customer and POS data to close the gap on missed sales and improve supply‑chain execution in an omni‑channel environment. He explains...
SK Telecom, Supermicro, Schneider Electric Partner on AI Data Centers
SK Telecom, Supermicro, and Schneider Electric signed an MOU at MWC26 to create a prefabricated, modular solution for artificial‑intelligence data centers. The joint effort will combine Supermicro’s high‑performance GPU servers, Schneider Electric’s MEP design expertise, and SK Telecom’s operational know‑how...
Building Support Systems for Digital Health Innovators
Digital health innovators require structured support systems to move from concept to clinical impact. The article outlines how regulatory navigation, technical mentorship, education, and institutional ecosystems collectively reduce risk and accelerate product readiness. Early integration of quality, compliance, and clinician...
Amazon Just Banned AI Agents From Its Seller Platform
Amazon updated its Business Solutions Agreement on March 4, 2026, adding a new Agent Policy that bans AI agents from accessing its seller platform unless they self‑identify and can be shut down on request. The agreement also forbids using Amazon’s...
Siemens to Deploy Cloud SCADA for Australian Renewables
Siemens, together with Mescada, will install an AI‑ready, cloud‑based SCADA platform for Global Power Generation Australia, covering eight renewable sites across four states. The Simatic WinCC Open Architecture system will handle roughly 300,000 data tags, delivering real‑time monitoring from a...
American Arbitration Association Launches Resolution Simulator, Expanding Its AI Arbitrator Tool
American Arbitration Association introduced the Resolution Simulator, an AI‑powered tool that generates simulated, non‑binding arbitration decisions for document‑only commercial and construction disputes. Building on its 2024 AI Arbitrator pilot, the platform analyzes uploaded pleadings, contracts, and evidence using large‑language models...
The People's Liberation Army's Perspectives on Artificial Intelligence Highlighting Integration as Key to "Intelligentization" Goals
China's People's Liberation Army (PLA) has placed artificial intelligence integration at the core of its “intelligentization” drive, a strategic priority set by President Xi Jinping. The PLA seeks to boost its kill chain, cross‑domain information sharing, multi‑agent collaboration, and data...
Phoebe Gates Raises $35M, Shuns Family Name, Focuses on Product
Phoebe Gates raised $35M for her AI startup Phia and refuses to use her last name in the pitch. She wants zero ties to privilege and took no money from her parents. Phia plugs into your browser and finds lower prices across...
API-First Insurance: When Brands Become Insurers
In this episode, Wayne Slavin, CEO and co‑founder of Sure, explains why the next biggest distributors of insurance will be non‑insurance brands that already own consumer relationships, and how Sure’s API‑first “digital insurance rails” enable these brands to embed frictionless,...
ANELLO Photonics & Mythos AI Partner to Secure GPS-Denied Maritime Navigation
ANELLO Photonics and Mythos AI announced a strategic partnership to deliver a plug‑and‑play maritime navigation system that operates without GPS. The solution merges ANELLO’s silicon‑photonic optical gyroscope (SiPhOG) inertial sensor with Mythos AI’s autonomous decision‑making software, creating an open‑systems INS...
Would You Trust an AI Pentester to Work Solo?
Security leaders face mounting pressure to outpace threats while accelerating AI adoption, yet only 36% are satisfied with current pentesting providers. AI‑powered pentesting promises unprecedented speed and scale, scanning massive codebases in minutes, but it falls short on contextual judgment,...
Defeating the Deepfake: Stopping Laptop Farms and Insider Threats
Cloudflare announced a partnership with Nametag to embed workforce identity verification into its Cloudflare One SASE platform, targeting the emerging "remote IT worker" fraud that leverages AI‑generated deepfake IDs and laptop farms. The integration uses OpenID Connect to require a...
OpenAI Updates ChatGPT to Cut ‘Cringe’ Replies and Improve Answer Quality
OpenAI rolled out the GPT‑5.3 Instant update for ChatGPT, aiming to curb the “cringe” tone that many users found overly therapeutic. The upgrade trims moralizing preambles, unnecessary caveats, and defensive refusals while sharpening conversational flow. OpenAI reports hallucinations drop 26.8%...
The New Media Stack & The AI Video Gold Rush
AI video platforms are democratizing content creation, allowing solo creators to produce cinema‑grade shows and brands to generate hundreds of ad variants at scale. Experts say speed and output will increase tenfold, but success will hinge on owning strong intellectual...
AI-Assisted Coding Needs More than Vibes; It Needs Containers and Sandboxes
In this episode, Docker President Mark Cavett discusses how containers are becoming essential for safely running AI‑generated code, emphasizing the need for hardened images to bridge the trust gap. He explains Docker’s new open‑source Docker Hardened Images (DHI) catalog, which...
AI Workloads Force a Fundamental Redesign of Middle East Datacentres
AI workloads are prompting a fundamental redesign of Middle East data centres, shifting from legacy digital architectures to AI‑centric designs. Huawei’s SuperPoD solution, announced at MWC 2026, delivers up to 96.6% UPS efficiency and a 25% smaller footprint to meet soaring...
AGTEK Launches Reveal Transform AI Module
AGTEK, a Hexagon subsidiary, announced Reveal Transform, an AI‑driven module that converts PDF construction plan sheets into CAD‑like linework and structured data. The tool automates text extraction, alignment, and classification, then exports the results to AGTEK’s Gradework platform for rapid...
How States and Localities Are Using AI
State and local governments are rapidly deploying generative AI to streamline core functions, from public‑defender video analysis to contract drafting and permitting. Tools like JusticeText have cut video review time, raising case clearance rates by roughly 8‑9 percent, while AI‑powered...
AI Recruiter Screens: What We Learned and Why We'll Keep Going
Zapier’s Talent Acquisition team piloted AI‑driven recruiter screens with Ezra AI Labs, processing roughly 250 interviews. The experiment cut screen time by 66%, from eight days to 2.75 days, and freed about 84 recruiter hours, equating to 5‑6 extra weekly...
How I Used Automation and AI to Redefine the EA Role
At Zapier, an executive assistant transformed a traditionally manual role by embedding automation and generative AI into daily workflows. By automating calendar triage, meeting prep, travel logistics and AI‑driven note‑taking, routine tasks shrank from 20‑30 minutes to under two minutes....
How PMOs Lead Responsible AI Transformation in Higher Education – with Ivonne Mejia
In this episode, Yvonne Mejia, Lead Project Portfolio Manager for the CSU Bay Region and PMI‑CPMAI certified AI champion, discusses how PMOs can responsibly drive AI transformation in higher education. She explains the CPM‑AI methodology, emphasizing the importance of starting...
OpenText Ramps up A/NZ Channel Investment
OpenText is ramping up its partner ecosystem in Australia and New Zealand to accelerate AI and cloud adoption after internally deploying its AI tools to more than 20,000 staff. The company will boost marketing, sales‑enablement and training resources, and is...
Grinn ReneSOM-V2H Module Runs Renesas RZ/V2H Vision AI Processor
Polish embedded‑systems firm Grinn unveiled the ReneSOM‑V2H, the world’s smallest System‑on‑Module built around Renesas’ RZ/V2H vision AI processor. The 37 × 42.6 mm board combines four Cortex‑A55 cores, two Cortex‑R8 real‑time cores, a Cortex‑M33 MCU and an 8 TOPS DRP‑AI3 accelerator for edge‑AI workloads....
Project‑specific Output Folders Streamline Context and Token Usage
In the spirit of keeping all project work self-contained, I modify the following structure by having an outputs folder for each project Instead of: ~/Claude-Workspace/ ├── context/ # Your standing context files ├── projects/ ...
One Connection Gives AI Agents Access to 200+ APIs
AI agents can do a lot. But if they can’t use real data, they end up guessing. @CaravoAI fixes that by connecting AI agents to 200+ APIs through one simple connection. That’s how automated workflows can scale → https://t.co/j6YKqRjWp1
ACAM Launches Australia’s Inaugural AI Readiness System for Marketers; THE ICONIC First to Use
The Australian Centre for AI in Marketing (ACAM) has unveiled AI‑Results Transformation (ART), the nation’s first end‑to‑end system for embedding AI into marketing performance. THE ICONIC became one of the inaugural adopters, using ART to create responsible, data‑driven customer experiences....
AI Agent Rebuilt macOS on Web, Defying Limits
I was surprised by how quickly Luca can add apps and functionality to this OS — it appears that the Agent built solid shared primitives and libraries and can layer in and maintain endless complexity.
Agentic AI Forces Unified ITOps‑SecOps for Resilience
Most orgs still treat ITOps and SecOps as separate universes, but incidents don't care about org charts. Agentic AI gives leaders a reason to redesign workflows around end-to-end resilience. #CIO #CISO #AI https://t.co/e3w3lXkvfc
Capgemini Joins OpenAI's Frontier Alliance to Scale Enterprise AI
Capgemini has joined OpenAI’s newly launched Frontier Alliance as a founding partner, creating a dedicated delivery function to scale AI agents for enterprises. The firm will deploy OpenAI‑certified professionals to tackle data readiness, integration, operating‑model design and governance challenges. Capgemini...
Max Fast Delivers Super‑intelligent, On‑demand Software Design
'@v0 Max Fast' (powered by Fast Opus) straight-up feels like frontend & design engineering super-intelligence. The quality, the speed, the consistency… It's costly, but it's the best taste of the future of generative, on-demand software you can get today. https://t.co/SNzS0vtFA9
Make Every UI Action API-Accessible for Agentic Work
everything in the GUI should be available in the API this will be the theme for us for the remainder of the year the UI becomes the preview layer for agentic work if you can click it, you should be able to prompt...
Solid Launches $20M Seed Financing Round
Solid, an enterprise software startup, announced its public launch alongside a $20 million seed financing round backed by Team8 and SignalFire. The capital will accelerate the company’s mission to make AI and analytics trustworthy by automating data semantics across organizations. Solid’s...
22% of Hong Kong Workers Use GenAI Daily, Compared to APAC Average of 29%
PwC’s 2025 Hong Kong Workforce Hopes and Fears Survey reveals that 61% of local employees are using AI at work, with 22% engaging with generative AI daily—below the 29% Asia‑Pacific average. The majority report productivity gains (77%) and higher work...
Did Alibaba Just Kneecap Its Powerful Qwen AI Team? Key Figures Depart in Wake of Latest Open Source Release
Alibaba released the Qwen3.5 small model series, praised for its high intelligence density and ability to run on consumer devices. Within 24 hours, technical architect Junyang "Justin" Lin and two teammates announced their departures, leaving the future of the Qwen...
Why AI Startups Are Selling the Same Equity at Two Different Prices
AI startups are increasingly using multi‑tiered valuations, selling the same equity at two distinct price points within a single funding round. Lead investors receive a discounted price while later participants pay a headline unicorn valuation, as seen in Aaru’s $450 million...
SaidText Brings Voice-Driven AI to Industrial Operations
SaidText is launching its voice‑driven AI platform in the United States, targeting industrial and manufacturing plants. The solution lets frontline workers record updates by speech, automatically transcribing and structuring the data for real‑time dashboards and analytics. Managers receive AI‑prioritized alerts,...
Seeking AI GRC Founder Connections After Impressive Vendor Meet
Looking to talk to more AI GRC platforms: 2 clients met with a vendor recently and were really impressed with the companies and space. Would love to meet some dope founders if anyone has recs
Microsoft: Securing AI Agents and Human Teams Crucial for Success
Microsoft’s inaugural Cyber Pulse AI Security Report reveals that over 80% of Fortune 500 firms already deploy low‑code AI agents, and the company forecasts 1.3 billion autonomous agents operating by 2028. Financial services account for roughly 11% of global agent activity, underscoring...
CHEO’s ThinkRare Algorithm Goes National
The CHEO Research Institute’s pediatric AI tool, ThinkRare, is moving beyond its home site to McMaster Children’s Hospital, Alberta Children’s Hospital and Stollery Children’s Hospital, marking one of Canada’s first national deployments of a real‑time, clinically embedded AI algorithm. ThinkRare...
Large Language Models Have the Potential to Level the Playing Field in Consumer Financial Complaints
A new research briefing examined 1.13 million US Consumer Financial Protection Bureau complaint narratives and found that large‑language‑model (LLM) assistance surged after ChatGPT’s release. The study shows LLM‑assisted complaints are more likely to secure consumer relief, with a measurable uplift in...
Aware Inc (AWRE) Q4 2025 Earnings Call Transcript
Aware Inc. posted fiscal 2025 revenue of $17.3 million, a marginal decline, while its net loss expanded to $5.9 million, reflecting higher technology and certification investments. The company secured ISO 30107 Level 3, ISO 27001, and FIDO2 certifications and earned best‑in‑class bias ratings in NIST...