AI News Archive: June 16, 2026 — Part 19
Sourced from 500+ daily AI sources, scored by relevance.
- Nvidia’s Jensen Huang says society needs ‘new social norms’ in the age of AI
Nvidia’s Jensen Huang says society needs ‘new social norms’ in the age of AI Boston Herald
- Finance and AI
Finance and AI Oxford Man Institute of Quantitative Finance
- ContraVault AI raises $3.1 Mn in pre-Series A round led by Chiratae Ventures
Bengaluru-based procurement intelligence startup ContraVault AI has raised $3.1 million in a Pre-Series A funding round led by Chiratae Ventures, with participation from existing investor Titan Capital Winners Fund. The fresh capital will be used to accelerate product development, strengthen the company’s AI capabilities, and support its expansion into international markets, particularly the United States. Founded in 2024 by Sayan Sen, Isha Juneja, and Tanmay Juneja, ContraVault AI offers an AI-powered platform that helps infrastructure and enterprise companies analyse tenders and requests for proposals (RFPs), identify risks, ensure compliance, and streamline bidding decisions. The startup focuses on digitising procurement and bid management processes, a segment that remains heavily reliant on manual workflows despite the scale of public and private infrastructure spending globally. According to the company, its platform has analysed more than one million tenders as part of its training dataset and is currently being used by some of India's largest infrastructure companies. The platform enables enterprises to discover opportunities, evaluate tender documents, assess commercial and compliance risks, accelerate go/no-go decisions, and prepare bid submissions with reduced manual effort. ContraVault AI has expanded beyond tender analysis to become a broader bid management platform serving infrastructure, construction, and energy companies. The startup plans to use the funding to enhance its domain-specific large language models (LLMs) and expand capabilities across procurement intelligence, workflow automation, and AI-assisted decision-making.
- Qualcomm in talks to buy Tenstorrent: Report
The talks were ongoing and the price could change, or the discussions could fall apart, according to the report. It was not clear if the price will include performance-based milestone payments, a structure used in the past to buy chip startups, the Information added.
- Microsoft Sued by Shareholders Over Expenses, Cloud Business, AI
Microsoft has been sued by shareholders who accused the company of defrauding them and inflating its stock price by failing to disclose slowing growth in its Azure cloud business and the need to spend billions of dollars on AI infrastructure. …
- US court rejects Musk's xAI suit accusing OpenAI of stealing trade secrets
The judge said asking job candidates to discuss their prior work was routine, and one could not infer that OpenAI pushed Li to leak anything confidential
- Sarvam AI turns unicorn; Razorpay confidentially files IPO papers
Sarvam AI turns unicorn; Razorpay confidentially files IPO papers YourStory.com
- Meta Introduces AI Mode on Facebook With AI-Powered Search for Public Posts, Reels
Meta has announced new AI-powered features for Facebook, led by the introduction of AI Mode, a search experience that generates answers using publicly available content from Facebook Groups, Reels and other Meta services. The update also adds AI-assisted creative tools, including collage templates, video montage effects and camera roll-based sharing suggestions. Users...
- Meta brings AI Mode to Facebook, drawing on public content across the platform
Meta brings AI Mode to Facebook, drawing on public content across the platform
- Meta’s New Facebook ‘AI Mode’ Pulls From Groups, Reels, and Posts
Meta AI Mode brings AI-generated answers to Facebook search using public posts, raising questions about reliability and source visibility. The post Meta’s New Facebook ‘AI Mode’ Pulls From Groups, Reels, and Posts appeared first on TechRepublic .
- Google's Next Pixel Drop to Reportedly Bring Screen Reactions, Gemini Omni Features to Pixel Phones
Google appears to be preparing its next Pixel Drop update for Pixel smartphones. While the Mountain View-based tech giant has yet to announce its arrival, a series of promotional videos have been spotted online, offering an early glimpse at what could be coming to eligible Google Pixel smartphones. Based on the teaser videos, Google could bring several AI-powered capa...
- Google’s June 2026 Pixel Drop arrives with floating app bubbles, screen reactions and many new AI tools
Google has started rolling out Android 17 and its June Pixel Drop update, which adds Bubbles, Screen Reactions, AI music creation, smarter calling features, and new Pixel Watch safety tools.
- AI’s next frontier, world models, and why China is ahead of the pack
They’ve been described as the “next frontier” of artificial intelligence (AI) – world models that go beyond large language models (LLMs) by simulating the physical environment. And in this emerging field, China has an edge over the United States. World models that understand the physical laws of the universe are already being used to train AI systems with embodied intelligence such as robots and self-driving vehicles. China has deployed such systems far more widely than the US – and this month,...
- Ant Revamps Alipay With AI Assistant in Biggest App Overhaul
Ant Revamps Alipay With AI Assistant in Biggest App Overhaul Caixin Global
- Alipay Unveils AI-Powered Revamp, Its Biggest Transformation in Two Decades
Ant Group is testing an AI-powered version of Alipay, representing the payment giant's most radical overhaul in 20 years — shifting from app-based services to conversation-driven interactions.
- Predicting LLM Safety Before Release by Simulating Deployment
Paper link Before releasing a new model, labs need to understand not just what it can do, but how it is likely to behave in real-world use, including where it might introduce new risks. This becomes even more important as capabilities increase. As part of our pre-deployment safety review, we leverage targeted evaluations, red-teaming, and other checks to understand model behavior. We’ve now started using a method for simulating model deployments before they happen, which adds a complementary signal: a deployment-like preview of how a candidate model may behave before it reaches users. Deployment Simulation is a method for simulating a future deployment before it happens. We do so by replaying previous conversations in a privacy-preserving manner with a new candidate model. By doing so, we can study how the new model responds in realistic contexts before release, including whether new undesired behaviors emerge and how often they may appear. In our GPT-5.4 study, these forecasts were informative. For categories whose production rates changed by at least 1.5x, deployment simulation predicted the direction of change 92% of the time, compared with 54% for a baseline built from challenging prompts. Simulated deployments also looked much closer to real production traffic on evaluation-awareness measures: traditional evals often visibly have stage lights; production prefixes mostly do not. The hardest case is agentic tool use, where realistic behavior depends on external state: filesystems, connectors, syscalls, network services, and prior tool results. We address this by using another model to simulate tool responses, with access to the original trajectory and time-matched codebase where possible. This is not a replacement for traditional evals, but it is a useful complement: safety evals should be forecasts with post-release scorecards, not just obstacle courses. We have already used insights from Deployment Simulation during model development to identify blind spots in traditional evaluations and inform mitigations and deployment decisions. As we make the pipeline easier to run, we expect it to play a larger role in the future model development process. Discuss
- Develop smarter AI agents with data fabrics
Develop smarter AI agents with data fabrics InfoWorld
- Building AI Agents for AR Glasses and XR Devices with NVIDIA XR AI
Developers building for AR glasses and wearable devices face an infrastructure gap. The hardware is ready, but creating AI experiences requires integrating live...
- NVIDIA Blackwell Tops MLPerf Training 6.0 with Industry-Leading Scale and Performance
NVIDIA delivered a clean sweep in MLPerf Training v6.0, the latest edition of industry-standard AI training benchmarks developed by the MLCommons consortium....
- HPE expands Private Cloud AI factory portfolio to support next-gen autonomous agents
Hewlett Packard Enterprise Co. today unveiled a massive expansion of its artificial intelligence infrastructure for private cloud deployments, enhancing its portfolio with Nvidia Corp.’s new Vera chips and tools for developers. The new offerings are meant to provide enterprises with everything they need to bring autonomous AI agents online within secure, on-premises environments to ensure […] The post HPE expands Private Cloud AI factory portfolio to support next-gen autonomous agents appeared first on SiliconANGLE .
- HPE AI Factory adds features to improve your experience with agents
HPE AI Factory adds features to improve your experience with agents IT Pro
- With Just 1 Click, Researchers Figured Out How to Hijack Microsoft Copilot to Steal Your Data
A recent Microsoft Copilot exploit demonstrates how AI can make existing cybersecurity bugs even more virulent.
- AI 에이전트는 늘어나는데 통제는 없다…CIO 덮친 거버넌스 공백
IT 리더들은 에이전트 시대의 거버넌스가 어떤 모습일지 미리 경험하고 있다. 그리고 그 모습은 결코 밝지 않다. IBM 기업가치연구소(IBM Institute for Business Value)가 CIO와 CTO를 대상으로 실시한 조사 에 따르면 응답자의 3분의 2는 자신이 완전히 통제하지 못하는 AI 시스템에 대한 책임을 지고 있다고 답했다. 직원과 각 사업부가 새로운 AI 에이전트를 자체적으로 구축·배포하고 있기 때문이다. 또한 응답한 IT 리더의 70%는 자사 조직이 IT 부서가 추적할 수 있는 속도보다 더 빠르게 기술 시스템을 도입하고 있다고 밝혔다. CIO들은 내년까지 조직 내 AI 에이전트 수가 38% 증가할 것으로 예상했지만, 예상되는 에이전트 확산 규모에 대비가 돼 있다고 답한 IT 리더는 10명 중 1명에 불과했다. IBM 기술혁신 CIO인 맷 라이테슨 은 이번 조사 결과가 AI 거버넌스에 대한 우려가 커지고 있음을 보여준다고 설명했다. 직원들이 IT 부서의 관여 없이 새로운 에이전트를 구축하면서 많은 IT 조직이 결과물, 보안, 비즈니스 가치 등을 추적하는 데 어려움을 겪고 있다는 것이다. 라이테슨은 “많은 기업이 더 많은 직원이 에이전트를 개발할 수 있도록 지원하는 정책을 운영하고 있다”라며 “이제는 IT 부서의 소수 인력만이 이런 솔루션을 개발하는 구조가 아니다”라고 말했다. IT 리더에게 닥친 위험 전문가들은 직원들이 AI를 자유롭게 실험하도록 장려하는 것은 분명 가치가 있지만, 동시에 IT 리더에게는 상당한 문제를 초래할 수 있다고 지적한다. 데이터 플랫폼 기업 도모(Domo)의 최고 AI·애널리틱스 책임자 벤 샤인 은 CIO와 CTO가 통제하지 못하는 AI 도구에 대한 책임까지 지게 되면서 조직 내 긴장이 커지고 있다고 분석했다. 그는 많은 조직에서 AI 도구가 IT 부서의 관리 체계에 편입되는 속도보다 더 빠르게 현업 부서에서 도입되고 있다고 설명했다. 샤인은 “문제는 AI가 무분별하게 배포되고 있다는 점이 아니다”라며 “AI 도입 속도가 거버넌스 체계가 적응하는 속도보다 훨씬 빠르다는 데 있다”라고 진단했다. 또한 상당수 AI 도입 사례는 직원들이 업무 효율성을 높이기 위해 합리적으로 판단한 결과이지만, CIO와 CTO는 이런 일이 벌어지고 있다는 사실조차 인지하지 못하는 경우가 많다고 설명했다. 그는 “마케팅 부서는 콘텐츠 제작 프로세스에 대규모언어모델(LLM)을 연결하고, 재무 부서는 예측 데이터를 챗GPT에 입력해 정리하며, 제품 부서는 새로운 에이전트에 고객 데이터셋 접근 권한을 부여한다”라며 “이런 활동이 조직 전체에서 동시에 이뤄져도 CIO에게는 보이지 않는다”라고 말했다. 샤인은 AI 거버넌스와 가시성이 현재 많은 조직에서 IT 리더가 우선적으로 해결해야 할 핵심 과제라고 평가했다. 그는 “AI가 실제로 무엇을 하고 있는지 볼 수 있는가”라고 반문하며 “AI 에이전트를 직원처럼 관리해야 한다. 누가 어떤 데이터를 사용했는지, 무엇을 생성했는지, 비용은 얼마나 발생했는지, 어떤 문제가 있었는지를 파악해야 한다”라고 설명했다. 이어 “AI 거버넌스는 선택적인 관리 체계가 아니라 기본 인프라”라며 “향후 24개월 동안 성공할 CIO는 데이터가 존재하는 동일한 계층에 가시성과 정책 집행 기능을 구축하는 사람이다. 사후적으로 별도의 AI 거버넌스 프로젝트를 덧붙이는 방식으로는 충분하지 않다”라고 말했다. 커지는 AI 거버넌스 공백 소프트웨어 테스트 기업 어플로즈(Applause)의 CTO 아티시 살비 도 AI 거버넌스 공백이 존재한다는 데 동의했다. 살비는 “모든 산업에서 에이전트 기반 워크플로와 AI, 다양한 제품이 매우 빠른 속도로 개발되고 있다”라며 “하지만 CIO나 CTO가 AI 시스템을 완전히 통제하지 못하는 상황에서 이를 어떻게 관리하고 통제하며 평가할지에 대한 이해는 개발 속도를 따라가지 못하고 있다”라고 설명했다. 그는 샤인과 마찬가지로 많은 조직이 IT 부서 외부 직원의 에이전트 배포를 장려하거나 허용하면서도 CIO나 CTO에게 이를 알리지 않는 점을 문제로 지적했다. 결과적으로 통제권은 에이전트를 구축한 사람에게 귀속된다. 살비는 “그 사람이 이를 관리하고 거버넌스를 수행하며 평가할 기술적 전문성을 갖췄을 수도 있고 그렇지 않을 수도 있다”라며 “누군가 업무 효율을 높이기 위해 무언가를 만들었지만, 그 대가로 토큰 비용과 보안, 규제 준수 등 다양한 문제가 발생할 수 있다”라고 말했다. 그는 많은 기업이 심각한 AI 거버넌스 과제에 직면해 있다고 진단했다. 살비는 “기업들이 수십 개의 에이전트 시스템을 구축하고 이를 조직 전반으로 확산시키면서도 제대로 된 평가 체계를 마련하지 못하고 있다”라며 “업무 속도를 크게 높이는 대신 직원들이 생산하는 결과물의 품질을 희생하고 있다는 사실을 인식하지 못하는 경우가 많다”라고 설명했다. 이어 “결국 이들은 평범한 수준의 결과물을 매우 빠르게 만들어내고 있으며, 결함까지 함께 생산하고 있을 가능성이 높다”라고 평가했다. 또한 에이전트 통제권이 IT 부서 밖에 있는 경우 위험은 더욱 커진다고 경고했다. 그는 “그 사람은 통제권을 적절하게 행사하는 데 필요한 도구와 노하우, 기술적 경험을 갖추지 못했을 수 있다”라고 말했다. 반면 데이터 보안 기업 마인드(MIND)의 공동 설립자 겸 CTO 이타이 슈워츠 는 IT 부서 외부에서 배포된 에이전트의 경우 아예 통제 체계 자체가 없는 사례도 많다고 주장했다. 적절한 가드레일이 마련되지 않으면 에이전트는 사실상 감독 없이 운영된다는 설명이다. 슈워츠는 “모든 AI 도구에는 책임을 지는 비즈니스 담당자가 있어야 하며 일반적으로 문서상으로는 그렇게 지정돼 있다”라며 “하지만 대부분의 시스템은 자율적으로 작동하고 결과가 항상 동일하지 않은 비결정론적 특성을 갖고 있다. 정해진 스크립트를 따르지 않기 때문에 실제로는 도구가 스스로를 통제하는 셈”이라고 설명했다. 슈워츠는 해결책이 AI 도입 속도를 늦추는 데 있는 것이 아니라 IT 리더에게 충분한 가시성과 정책 집행 도구를 제공하는 데 있다고 강조했다. 그는 “내가 만난 어떤 기술 리더도 AI 도입 속도를 늦추길 원하지 않는다”라며 “그들은 빠르게 혁신하면서도 안전성을 확보할 수 있는 더 나은 도구를 원한다. 필요한 것은 더 많은 경계가 아니라 더 나은 도구”라고 말했다. 통제된 AI 플랫폼 구축 라이테슨은 IBM이 하나의 해법으로 통제된 환경에서 직원들이 자체 AI 도구를 개발할 수 있는 에이전트 플랫폼을 구축했다고 소개했다. 새롭게 생성되는 에이전트는 보안과 개인정보 보호 기준을 충족하는지 점검받으며, 기존 IBM 도구와 기능이 중복되는지도 검토된다. 라이테슨은 “우리는 매우 초기 단계부터 엔터프라이즈 플랫폼을 구축하고 직원들이 그 안에서 솔루션을 개발하도록 했다”라며 “신뢰할 수 있다고 판단한 특정 모델만 사용하고 데이터를 보호함으로써 CISO와도 안심하고 협력할 수 있는 환경을 마련했다”라고 설명했다. 이어 “해당 시스템이 실제로 무엇을 하는지 알고 싶다”라며 “개인의 생산성을 높이는 기능인지, 전체 업무 프로세스를 지원하는 것인지 파악해야 하며, 조직에 어떤 비용이 발생하고 어떤 가치를 창출하는지도 확인할 수 있어야 한다”라고 말했다. dl-ciokorea@foundryco.com
- Qualcomm launches Snapdragon Reality Elite and a white-label toolkit for AI glasses, betting the next platform is not a phone
Qualcomm announced two products on Tuesday aimed at positioning the company as the silicon supplier for whatever computing device eventually displaces the smartphone. The first is Snapdragon Reality Elite, a mixed reality chip platform with substantially improved AI processing for headsets and tethered glasses. The second is START, a white-label toolkit that gives eyewear manufacturers […] This story continues at The Next Web
- Qualcomm takes spatial computing into the AI era with Snapdragon Reality Elite
Qualcomm Technologies Inc. today unveiled a brand-new chip featuring the Snapdragon Reality Elite, a rebrand of its existing line of powerful silicon designed to power next-generation immersive virtual reality and mixed reality experiences. Reality Elite is the successor to the XR2+ Gen 2, announced during the Augmented World Expo in 2024, representing a generational leap […] The post Qualcomm takes spatial computing into the AI era with Snapdragon Reality Elite appeared first on SiliconANGLE .
- Study finds $18 trillion in enterprise value at risk as AI readiness challenges mount: Genpact Research
A new study by Genpact and HFS Research estimates that Global 2000 companies could unlock nearly $18 trillion in value by addressing longstanding operational challenges that continue to limit the […] The post Study finds $18 trillion in enterprise value at risk as AI readiness challenges mount: Genpact Research appeared first on Express Computer .
- This 25-Year-Old's Sleep Problems Gave Him Vertigo. His Response? An AI Mattress Cover That Hit Eight Figures Within Three Months of Launch
This 25-Year-Old's Sleep Problems Gave Him Vertigo. His Response? An AI Mattress Cover That Hit Eight Figures Within Three Months of Launch entrepreneur.com
- Apple Is Reportedly Planning AirPods With Cameras for 2027
The company's next major product wave could include AI-focused AirPods, smart glasses, a 20th-anniversary iPhone and a new foldable.
- Infosys Collaborates with Valmet to Reimagine IT Operations with AI Accelerated Transformation
Infosys today announced a strategic, long-term collaboration with Valmet, a global technology leader in serving process industries. As a part of this engagement, Infosys will modernize Valmet’s core IT services and deliver end-to-end IT transformation to enhance operational efficiency and enable closer alignment between IT operations and business priorities. Through this collaboration, Infosys will focus on leveraging […] The post Infosys Collaborates with Valmet to Reimagine IT Operations with AI Accelerated Transformation appeared first on CXOToday.com .
- G7 leaders discuss 'trusted partners' access to cutting-edge US AI models, sources say
G7 leaders discuss 'trusted partners' access to cutting-edge US AI models, sources say The Straits Times
- G-7 leaders discuss ‘trusted partners’ access to cutting-edge US AI models: Sources
G-7 leaders discuss ‘trusted partners’ access to cutting-edge US AI models: Sources The Straits Times
- These companies are finding new ways to use AI and tech for security, health, and more
Even as the AI revolution moves toward agentic tools, there are still hosts of uses for more classical machine-learning and computing—particularly in healthcare. That’s where many of the AI and technology honorees of this year’s World Changing Ideas are operating—like MEandMine and Expert Radiology—but not all of them. Others tackle the need to identify AI-created content (Copyleaks), explore the future of in-home robots (Humanoid), and focus on the future of clean energy (Fuze Energy). Read about all of them below. Winners AI Logic GenAI detection , Copyleaks Speculating on whether someone used AI for writing can be fun and games, but for many professionals (and students submitting papers), it can be a serious accusation. AI Logic, a program from Copyleaks to detect AI-generated writing, is designed to make it clear whether or not a piece of text is written by AI—but more importantly, explain why and how confident it is. The software identifies commonly used AI phrases and checks if any part of the text is from other known AI-generated sources. It’s now integrated into many educational systems, giving teachers an easy way to monitor students’ work, and the company say it is processing 650,000 AI checks a day. AI-powered therapy support , Grow Therapy Grow, an online therapy platform with more than $1 billion in revenue, launched its AI notetaker to cut its therapists’ recordkeeping time so they can focus on connection with patients. The new software captures the visit so therapists can approve notes after instead of focusing on them during the session, reducing providers’ time spent on documentation by 70%. The AI service also helps patients complete journaling between sessions—something more than 35,000 Grow patients do as part of their mental health work—which can generate quick summaries to serve as a starting point for the next session. Fighting cyberattacks with specialized AI , Darktrace Digest is an AI model that’s been trained on more than a million cybersecurity incidents. It’s one of the features of Darktrace Cyber AI Analyst, a service to detect breaches quickly—and stop them before they cause disruption or result in the exposure of sensitive data. Working with corporate and government clients, the software, which autonomously investigates alerts to allow human security professionals to prioritize their focus, conducted more than 90 million investigations of cybercrime in 2024. At one client, a water district in California, Darktrace’s software led to reduced triage time for every security alert from three hours to 20 minutes. Gamified children’s mental health assessment , MEandMine When students play the games on MEandMine, it may seem like fun, but it’s actually helping them build self-regulation—and gives teachers insight into early risk signals of disruptive behavior. Daily check-ins about emotions, energy, social connection, and physical health help the company map a child’s “regulation spectrum,” then the company’s AI engine analyzes over 160 behavioral signals in the 200-plus games offered on the platform. In the more than 200 schools and health centers using MEandMine (serving over 40,000 students), there’s been a reported 70% increase in learning readiness and a 40% improvement in behavioral health metrics. Healthcare intelligence platform, Ambience Healthcare When you’re sick, you want your doctor focused on your illness—not admin work or billing codes. Ambience Healthcare is an AI-based technology that listens to patient conversations and handles the forms, billing, and other business-side tasks for doctors after the appointment, so the doctor can focus on the patient. In tests, Ambient resulted in a 41% reduction in active documentation time and a 39% reduction in documentation completed after hours for its users. With training to make it viable for more than 200 different medical specialties, the service also offers summaries of’ charts, giving doctors helpful summaries of patients’ medical histories before visits. After a competition with a number of other AI healthcare services at the Cleveland Clinic, Ambient Healthcare was selected for a five-year partnership. PrecisionPlus V3 , Expert Radiology Many AI companies have attempted to read radiology results better than radiologists. PrecisionPlus V3 is trying, instead, to make their findings more accessible. After an MRI, the software automatically enhances the report with helpful illustrations, annotated images, and a concise AI-generated summary written in layperson’s language. This makes the report a helpful document for both the patient (who now better understands their results) and their clinician (who doesn’t need to call the radiologist for extra information). PrecisionPlus V3 is already in use by Expert Radiology, a teleradiology network operating in all 50 states that covers more than 150,000 cases a year. Honorable Mentions FuZE-3 , Zap Energy Humanoid , Humanoid Explore the full list of Fast Company’s World Changing Ideas , 191 projects that are making the world more accessible, equitable, and sustainable.
- These projects help companies leverage AI to tackle pressing business needs
Artificial intelligence is the common theme that connects the winners in the Business Products and Services category of 2026 World Changing Ideas. AI proved its dominance in this year’s innovations for businesses, helping to accelerate critical clinical trials, fight online hackers, and even rival LLMs at their own search game. Winners Autonomous offensive cybersecurity , Xbow For all its benefits, AI is also supercharging cyberattacks, 80% of which use the technology for everything from data breaches to service outages. Xbow’s response: fight AI with AI. The cybersecurity company equips businesses with its AI-enabled cybersecurity platform, which simulates an AI hacker then identifies vulnerabilities in systems that could be ripe for real AI hacks. In 2025, it climbed up the ranks and became the number-one ethical hacker tool in the world. With this tool taking on continuous, proactive scanning, human cybersecurity professionals can spend more time on reactive defense. Xbow also sees its cybersecurity tool as democratizing access for smaller companies, startups, nonprofits, and the public sector, who have previously been priced out of top-tier security. Dewey AI Site Search , Dewey Labs As internet searches shift to tools like ChatGPT and Google Gemini, users are awash in information, and LLMs can easily hallucinate, confidently providing incorrect information. Many websites act as their own mini repositories of specific information, but the search function on most is lackluster. That’s where Dewey Labs’ AI site search comes in. It helps site owners lean into the expert knowledge that they have amassed, making it easier for users to find what they need. Through a pilot with ParentData and trusted economist Emily Oster, Dewey helped the sites’ users achieve a 91% success rate in finding what they wanted (compared with 28% for traditional site search). That pilot turned into a full launch in 2025 that has brought it to more sites, including Spotlight PA for election information. With 1.6 million answers provided, the company is eyeing ways to use the tool for nonprofits and public health sites. Manifold.AI platform , Manifold.AI The sluggish pace of clinical trials for drug discovery—due to legacy systems that function with Excel and manual data—can often severely delay progress. That results in the loss of valuable time for patients as well as a waste of time for researchers and an estimated $80 billion in annual inefficiency. While solutions usually address isolated points of the process, Manifold.AI is overhauling the entire system, using AI agents to orchestrate the entire clinical trial process, from trial design to execution and analysis. Last year, the tool moved from concept to product, and it has already been used by major institutions, including the Broad Institute of MIT and Harvard, which collaborated to build a new AI life sciences research platform. Meanwhile, the American Cancer Society is using Manifest to handle one of the largest population studies of Black women in the U.S. Honorable Mentions U.S. Hub for Rare Earth Magnet Recycling , Cyclic Materials Yseop Copilot , Yseop Explore the full list of Fast Company’s World Changing Ideas , 191 projects that are making the world more accessible, equitable, and sustainable.
- AI to be used to identify plants at risk of extinction
AI has been identified as a potential turning point in the "race against extinction" faced by botanists trying to save plants.
- ChatGPT Knows Too Much: 8 Ways to Lock Down Your Privacy
ChatGPT Knows Too Much: 8 Ways to Lock Down Your Privacy PCMag Australia
- ChatGPT Knows Too Much: 8 Ways to Lock Down Your Privacy
ChatGPT Knows Too Much: 8 Ways to Lock Down Your Privacy PCMag UK
- Databricks releases agentic coworker Genie One
Databricks today at its Data & AI Summit announced the release of Genie One, a new agentic coworker designed for business teams. This product launch expands the Genie suite of AI tools, which now includes Genie One, Genie Agents, and Genie App Builder. The Genie Suite Components Genie One functions as an automated coworker that … continue reading The post Databricks releases agentic coworker Genie One appeared first on SD Times .
- SiMa Launches Agentic Development Environment for Physical AI
Edge AI chip company said it can speed up engineer productivity when transferring to its hardware from months to hours. The post SiMa Launches Agentic Development Environment for Physical AI appeared first on EE Times .
- HPE unveils a raft of new networking products for AI workloads at Discover 2026
HPE unveils a raft of new networking products for AI workloads at Discover 2026 IT Pro
- Salesforce Makes $3.6B Play for AI Customer Service
Salesforce will buy Fin, formerly Intercom, for about $3.6 billion to expand Agentforce with AI customer service agents and support automation. The post Salesforce Makes $3.6B Play for AI Customer Service appeared first on TechRepublic .
- OpenAI’s ChatGPT to debut on GenAI.mil in ‘early July’
The company is working closely with the Pentagon’s Chief Digital Artificial Intelligence Office to deliver more model access via GenAI.mil.
- US counterintelligence agency looks to AI to accelerate background checks
A Defense Counterintelligence and Security Agency official says advanced AI can cut parts of the vetting process from months to hours.
- French spy agency drops US AI giant Palantir
PARIS - France’s government took a new step on Tuesday in search of “strategic autonomy” for its intelligence services, announcing that it would ditch the data-sifting systems of the American AI giant Palantir in favour of a domestic provider.
- Nvidia's Huang pledges AI will boost manufacturing jobs. A test will come in Texas
Nvidia's Huang pledges AI will boost manufacturing jobs. A test will come in Texas San Francisco Chronicle
- Nvidia’s Huang pledges AI will boost manufacturing jobs. A test will come in Texas
Nvidia’s Huang pledges AI will boost manufacturing jobs. A test will come in Texas Boston Herald
- Why This AI Data Center Stock Is Surging on an AMD Partnership
Why This AI Data Center Stock Is Surging on an AMD Partnership Barron's
- Xreal's Android XR Aura Glasses Are Coming This Fall, With a Brand New Qualcomm Chip Inside
Qualcomm's new Snapdragon Reality Elite could show up in a lot more future headsets, but it's appearing in Google and Xreal's upcoming Android XR glasses first. It's promising significant boosts for graphics, battery and AI.
- Checkmarx Adds Hybrid SAST Engine to Improve AppSec in AI Era
Checkmarx Adds Hybrid SAST Engine to Improve AppSec in AI Era DevOps.com
- Can AI help us age better? Bay Area scientists are trying to find out
Can AI help us age better? Bay Area scientists are trying to find out East Bay Times
- AI can clone your voice. Why that’s powerful — and dangerous
AI can clone your voice. Why that’s powerful — and dangerous EurekAlert!