Know What's Happening in AI
Know What's Happening in AI
OpenAI faces Florida probe as it rolls out $100‑per‑month Pro tier
OpenAI is under investigation by the Florida Attorney General after allegations that ChatGPT was used to plan a university shooting. The company pledged full cooperation and highlighted ongoing safety upgrades for its large user base. Simultaneously, OpenAI launched a $100‑per‑month Pro subscription for developers using its Codex coding tools.
Also developing:
By the numbers: Optasia secures $330M debt facility
Liquid AI unveiled LFM2-24B-A2B, a 24‑billion‑parameter sparse Mixture‑of‑Experts model, paired with the open‑source LocalCowork desktop agent. The architecture activates only about 2 billion parameters per token, fitting into a ~14.5 GB RAM footprint on an Apple M4 Max. LocalCowork runs entirely offline, using the Model Context Protocol to call 75 pre‑built tools while logging actions locally. Benchmarks show ~385 ms latency per tool selection and 80% accuracy on single‑step tasks, though multi‑step chains succeed only 26% of the time.
What does @OpenAI GPT 5.4 (just released this morning) mean for developers? You talked about it. My AI agents read it. Wrote a report. Sent it over to @NotebookLM. Which made this video. Using the new cinematic quality released yesterday. This was...
HubSpot’s 2026 automation suite centers on the Breeze AI Assistant, which creates workflows and specialized agents from natural‑language commands. The platform separates high‑volume marketing automation via Workflows from one‑to‑one sales automation through Sequences, while AI agents handle prospect research, data...
Ericsson and Australian carrier Optus have launched a joint trial of an artificial‑intelligence model that predicts whether a device on one 5G Standalone frequency layer is also covered by another. The model, trained on months of live Optus data, can...
The NeuroGrid competition, hosted on Hack The Box, pitted 120 AI‑augmented teams against 958 human‑only teams across 36 security challenges. AI‑augmented participants completed 73% of challenges, far outpacing the 46% completion rate of human‑only teams. The advantage was strongest on...
Thyme Care has launched an Integrated Social Support (ISS) model that places licensed master‑level social workers as the first point of contact for cancer patients. The model proactively screens 100% of members at enrollment, uncovering that 53% face financial toxicity,...
An upcoming webinar on 15 April 2026 will examine how agentic AI is reshaping banking. It will showcase use cases such as real‑time fraud prevention, predictive credit‑risk analytics, and back‑office automation, while addressing ethical, privacy, and regulatory concerns. Speakers will...
I've ran OpenClaw for over a month now I've had it in a group chat with 26 friends who all played with it, tried to hack it, made a pretty cool game with it which it kept self improving called...
KPMG partner Harnath Babu warns that most agentic AI projects stall between proof‑of‑concept and production because organizations treat pilots as deployments. He outlines a four‑step playbook—prove value, harden the platform, institutionalize governance, and scale autonomy—to move agents from isolated tasks...
Nigerian‑born entrepreneur Tolu Fagbola, together with his US‑based brother, launched Vimedra to apply voice‑AI and predictive analytics to US healthcare. Their patent‑pending algorithm fuses social determinants with clinical data, creating risk profiles that flag patients likely to generate costly emergency...
The AI tool I called “best in class” just got dethroned - in three months. So when health systems go all-in on one AI vendor, is that a BIG mistake? I was loyal to ChatGPT. Then I became Gemini-pilled. Recently my...
Microsoft announced it will continue offering Anthropic’s AI models across its consumer and enterprise platforms, including Microsoft 365, GitHub Copilot, and the AI Foundry, while barring only the Department of War. The Pentagon has labeled Anthropic a supply‑chain risk, prompting the...
Eurovision Sport, the EBU’s free‑to‑air streaming service, has teamed with AI specialist CAMB.AI to deliver live and on‑demand subtitling for the Milano Cortina 2026 Paralympic Winter Games. The partnership will provide real‑time, contextual speech‑to‑text transcription across all six Paralympic sports, preserving editorial...
Validio, a Stockholm‑based data‑quality automation startup, secured $30 million in Series A funding, bringing its total capital to $47 million after an 800 % ARR surge last year. The round was led by Plural with participation from Lakestar, J12 and several angels. Validio’s AI‑driven...
Today, @DrJessePines, writing in @Forbes, covered the NOHARM benchmark — one of the most rigorous evaluations of clinical AI to date. 100 real medical cases across 10 specialties, 12,747 specialist annotations on beneficial and harmful medical actions, and 31 AI systems rigorously...
AI voice startup Elevenlabs has appointed former Conversant Media executive Michael Costa as its go‑to‑market director for Australia and New Zealand. Costa, who has held senior roles at Salesforce, Twilio and Mixpanel, will lead the company’s expansion despite still being based...
Sydney‑based Lumonus secured an additional $3 million, raising its Series B to $28 million. The AI‑powered platform automates radiation‑oncology workflows, already supporting over 280,000 cancer treatments and 75,000 treatment plans across the US, Australia and Europe. Funds will expand U.S. go‑to‑market teams, boost...
A lot of people think AI will replace engineers. But AI still needs: 🩷 infrastructure 🩷 pipelines 🩷 monitoring 🩷 security 🩷 deployment systems ✨ AI runs on top of DevOps. Not instead of it ✨
Who’s Accountable when AI Agents Act “Agency isn’t a feature—it’s a transfer of decision rights,” says McKinsey Partner Rich Isenberg. “The question shifts from ‘Is the model accurate?’ to ‘Who’s accountable when the system acts?’” https://t.co/Lgqcxr3zq4 @McKinsey #AIGovernance
NCTC’s AI Center for Excellence is wrapping up four pilot projects with three member companies, with results expected by May. The pilots aim to prove ROI for use cases such as network‑outage prediction and churn reduction. Successful outcomes could move...
X Defunds The Deepfakes: Undisclosed AI War Clips To Face 90-Day Revenue Ban | Republic World https://t.co/hUVPj3L7Nj #war #wardeepfake #Deepfake #DeepfakeDetection #AIGeneratedContent #AISecurity #DigitalForensics #SyntheticMedia #MediaManipulation #CyberSecurityAwareness #Disinformation #Misinformation #AIEthics #AIRegulation #TrustInMedia #ContentAuthentication #OnlineSafety
Overheard in Silicon Valley: “The current brainiac position is that AI *should* dictate war policy to the government but *should not* answer questions about your hangnail.”
OpenAI’s ChatGPT has entered its fourth week of serving contextual ads, with Sensor Tower reporting that retail and grocery promotions dominate the ad inventory. Over a two‑week sample, more than 100 distinct brand ads appeared, outpacing other categories such as...
Forrest says AI the stuff of nightmares, but also critical to Fortescue’s push to real zero #energysky -- via Renew Economy: https://t.co/CtSKQsQa1P https://t.co/B2sbBHDjw9
Contributor Spotlight: Shantae Hansen (p. 34) — as AI gets embedded into MES/ERP + IIoT + plant networks, cybersecurity has to be “built-in,” not bolted on. Practical moves: segment OT/IT, least-privilege + MFA, disciplined patching and tested offline backups. #Cybersecurity https://t.co/zvLr3eVkMd
DoubleVerify’s Fraud Lab has uncovered AutoBait, a coordinated network of more than 200 AI‑generated lifestyle sites that churn out clickbait articles and hyper‑realistic images to harvest advertising revenue. The operation delivers tens of millions of ad impressions each month, with...
What good things is AI doing? Here's 10 things my AI agents found here on X that are examples of good things AI has done for humans. +++++ 1. "ChatGPT saved my life" — grocery store conversation about an ultra-rare diagnosis "Doctors...
Customers.AI is an identity‑resolution and deliverability platform built for Shopify brands that run Klaviyo. Its X‑Ray pixel lifts visitor identification from roughly 15‑20% to 60‑70%, feeding anonymous shoppers into Klaviyo profiles. The Alfred AI agent scores these visitors daily, enabling...
Salesforce has launched three Slack‑native features—Slackbot, Activity Timeline, and Agentforce Sales—to embed CRM data directly into Slack conversations. Slackbot delivers AI‑driven answers from Salesforce records, Activity Timeline streams emails, calls and tasks into the chat view, and Agentforce autonomously monitors...
Enterprise Security Operations Centers face overwhelming alert volumes—averaging 4,484 alerts daily, with 67% never investigated. Autonomous triage platforms such as D3 Security's Morpheus use AI to ingest, investigate, and respond to 100% of alerts, cutting false‑positive handling by 99% and...
There was a nice time where researchers talked about various ideas quite openly on twitter. (before they disappeared into the gold mines :)). My guess is that you can get quite far even in the current paradigm by introducing a number...
At the 2026 Committee of Supply, Singapore pledged SG$800 million to launch a semiconductor RIE Flagship, SG$60 million for a National Semiconductor Translation and Innovation Centre, and another SG$800 million for a Decarbonisation Grand Challenge. A*STAR will spearhead these programmes alongside the Economic...
The Subang Jaya City Council has launched SENSA, an AI‑driven pedestrian safety system, at a school crossing in the city. The solution combines analytical cameras, sensors, audio warnings and blue LED lights to alert drivers and pedestrians in real time....
Harvard’s Kempner Institute has opened its 2026 Accelerator Awards request for proposals, offering faculty access to the Kempner AI Cluster. Winners receive up to 64 GPUs for a 30‑day period to conduct advanced intelligence research. The program, now in its...
this is gonna get no views vs the confirmation bias machine, but pls remind everyone that the @OfirPress literal swebench author doesnt endorse this cheap sample benchmark and you need to run about 30-60x compute that margin labs is doing...
The U.S. Department of Labor unveiled an AI Literacy Framework that defines five core competency areas for workers to use and evaluate generative AI responsibly. The framework outlines understanding AI principles, exploring real‑world uses, directing AI effectively, evaluating outputs, and...
In this episode the hosts discuss Canva's recent acquisitions of Cavalry, a UK‑based 2D motion‑animation studio, and Mango AI, a tool for optimizing ad performance, highlighting how these moves expand Canva's capabilities into animation and AI‑driven marketing. They note Canva’s...
FounderCoHo is hosting a free, invitation‑only after‑hours side event during NVIDIA GTC 2026 in Sunnyvale. The two‑hour program focuses on moving from frontier models to production‑ready agents and on scaling AI distribution through open‑stack infrastructure. Speakers include senior leaders from...
COSO released "Achieving Effective Internal Control Over Generative AI (GenAI)", a guidance that aligns its Internal Control‑Integrated Framework with the unique risks of generative AI. The publication translates the five COSO components into concrete controls for eight GenAI capability types...
Healthcare AI startups are moving from traditional SaaS subscriptions to transaction‑based pricing, charging per successfully completed task such as scheduling or intake workflows. Companies like Prosper AI, MedCalls.ai and Intelliclinic tie fees to usage, promising alignment with customer ROI. Investors...
Note that "exposure" here does not necessarily mean "job loss". It means AI can automate -- or is now automating -- many of the TASKS involved in a job. That CAN manifest as job loss, but it can also free...
OpenAI is eons behind here, esp for financial modeling. Think they were in early on Rogo which may help (probably good acquisition opportunity tbh) but Anthropic has been crushing and Mistral is doing well with EU banks (mainly because of...
SCSP released an interactive report estimating that artificial intelligence could influence 25 % to 64 % of tasks across all 131 Army officer MOS, with combat arms still seeing over a quarter of duties affected, especially during deployments. The study proposes four...
"Trust, but verify" is the new data mantra for AI literacy. Teach employees to validate LLM and AI agent outputs instead of treating them as unquestioned truth. #DataCulture #AI #CIO https://t.co/aiB5P99ido
Alex has been tracking the evidence on whether AI is showing up in the productivity stats. Seems like that may have actually started.
EY announced the launch of EY.ai Agentic for Sales, an AI‑driven sales orchestration platform built with Snowflake’s data cloud and Canva’s design automation. The solution integrates real‑time data intelligence, automated content generation, and EY’s consulting frameworks to streamline prospecting, outreach,...
In this episode the host showcases Bixby’s deep OS integration on Samsung Galaxy devices, demonstrating rapid split‑screen multitasking, seamless YouTube playback without Premium, and a macro‑tracking workflow that pulls nutrition data from a food‑ordering app. The host argues that Bixby...
Don’t sleep on this. No matter how many egos are involved money talks and the reality is that no other lab is able to deliver what Anthropic provides right now.
Google Cloud announced a suite of Gemini‑powered AI agents at HIMSS26, partnering with major health players including Humana, CVS Health, Highmark, Waystar and Quest Diagnostics. The agents target patient engagement, workflow automation, revenue‑cycle management and lab‑result interpretation, leveraging the Vertex...
True Positive Weekly #151 curates the latest AI breakthroughs, from Google’s efforts to teach large language models Bayesian reasoning and map reading, to an open‑source genome model trained on trillions of DNA bases. The issue also spotlights ultra‑compact models, such...
