This Week in NLP #378
Keep up with what happened in Natural Language Processing in the week ending Friday 13th February 2026.
Above the Fold
Overwhelmed? Here’s our pick for five things to know about this week.
Alphabet is raising US$185B through a seven-part bond offering to fund AI infrastructure and capital expenditures through 2026. [Yahoo Finance]
Apple is delaying the rollout of enhanced Siri features in iOS 26.4 due to testing challenges with processing queries and voice control. [Yahoo Finance]
Google’s Gemini 3 Deep Think update achieves breakthrough reasoning capabilities, excelling in scientific, coding, and logical tasks with unprecedented AI performance. [Marktechpost Media]
OpenAI has begun testing ads in ChatGPT for US free-tier users, promising privacy and claiming to make AI accessible to billions who can’t pay for subscriptions. [The Register]
Anthropic’s Super Bowl ads mocked ChatGPT’s ad-supported tier, prompting Sam Altman to criticize the company for what he perceived as dishonest messaging. [TechRadar]
Now read on for everything else that happened in NLP this week.
Making News:
What makes this newsletter different?
This Week’s Topics:
If you’re reading this in a desktop web browser, you can access a navigation menu at the left-hand edge of this window.
The Generative AI Wars
Alphabet has issued a rare 100-year bond and raised US$20B in debt, signaling Big Tech’s massive capital needs for AI infrastructure investments. [European Business Magazine]
Anthropic has raised US$30B in a funding round, valuing the AI startup at US$380B. [Yahoo Finance]
Anthropic claims Claude is now writing its own code, with internal teams using the AI to speed up development and improve security. [ITPro]
And Anthropic launched Claude Cowork for Windows, expanding its AI agent software to 70% of desktop computing and intensifying its partnership with Microsoft. [VentureBeat]
Apple’s capital expenditure has sharply declined, contrasting with peers’ massive AI infrastructure investments, raising questions about its competitive strategy in the emerging AI landscape. [WebProNews]
Google is exploring AI-powered advertising strategies, experimenting with ads in AI Mode while maintaining organic search results and avoiding direct integration in Gemini. [The Register]
Meta AI is preparing to release new Avocado models, expand features with app integrations, browser and voice agents, and potentially introduce a multi-model capability called Big Brain. [TestingCatalog]
SemiAnalysis president Doug O’Laughlin criticized Microsoft’s AI strategy, claiming the company is falling behind rivals despite its OpenAI partnership. [Yahoo Finance]
Microsoft is developing its own advanced AI models to reduce dependence on OpenAI, investing heavily in infrastructure and alternative AI partnerships. [Yahoo Finance]
Apollo Global Management is nearing a US$3.4B loan to an investment vehicle planning to buy Nvidia chips for Elon Musk’s xAI. [Yahoo Finance]
OpenAI is experiencing strong growth in ChatGPT and Codex, preparing to launch a new chat model while seeking a potential US$100B funding round. [CNBC]
OpenAI’s Codex app reached 1 million downloads in its first week, signaling growing demand for AI-powered coding tools. [VentureBeat]
Snowflake is accelerating innovation through strategic AI partnerships, particularly with OpenAI, to help enterprises leverage their data for meaningful AI insights. [ITPro]
SoftBank is expected to post a healthy profit from its OpenAI investment, with potential additional funding and growing market interest. [Yahoo Finance]
AI Supremacy
Huawei could potentially revive its chip technology if US sanctions are relaxed, according to Intel CEO Lip-Bu Tan. [The News International]
India is hosting the AI Impact Summit to champion the Global South’s voice and democratize AI resources, while navigating complex technological, economic, and geopolitical challenges. [Rest of World]
Nvidia has received limited licenses to sell AI chips to China, subject to strict US government-imposed restrictions and revenue-sharing terms. [Yahoo Finance]
Ro Khanna, a US House Democrat, signalled openness to selling older Nvidia AI chips to China while opposing sales of newer generations. [Yahoo Finance]
Taiwan’s Vice Premier Cheng Li-chiun rejected the US demand to move 40% of chip production to America, emphasizing that advanced semiconductor technology will remain on the island. [Asia Financial]
Sovereign AI
A Gartner report predicts global sovereign cloud spend will surge 35.6% to US$80B in 2026, driven by geopolitical tensions and data sovereignty concerns. [CIO Dive]
Gartner predicts European sovereign Cloud IaaS spending will more than triple from US$6.7B in 2025 to US$23.1B in 2027. [TechRadar]
Dutch banks ABN Amro, ING, and Rabobank are seeking European alternatives to US technology to enhance digital sovereignty and reduce geopolitical risks. [NL Times]
The EU is planning a sovereign military data-sharing platform by 2030, aiming to reduce dependence on US tech and avoid potential ‘kill switch’ risks in military infrastructure. [BankInfoSecurity]
And the EU has inaugurated NanoIC, a €700m semiconductor pilot line at the research hub imec in Leuven, aimed at advancing chip technologies and strengthening Europe’s global semiconductor position. [The Next Web]
But EU capitals say deleting US tech is not realistic. [Politico]
Datacenter consultancy BCS reports only 20 percent of European datacenters are AI-ready, with skills shortages and infrastructure challenges threatening future capacity expansion. [The Register]
Fractile, a UK chip startup, pledges £100m investment, with AI minister Narayan urging bold risks to establish Britain as an AI superpower. [Tech.eu]
A UK Treasury Select Committee report warns UK financial services firms are overly dependent on a small number of US tech companies for AI and cloud services. [Money Marketing]
Aramco and Microsoft signed an MoU to accelerate industrial AI deployment, strengthen digital capabilities, and support Saudi Arabia’s technological sovereignty. [Yahoo Finance]
Canada’s innovation ministry released an AI strategy consultation report with broad recommendations, using AI tools to analyze over 11,000 submissions. [BetaKit]
India and the United States outlined a framework to expand trade in advanced technology products, including GPUs, as part of a broader strategic cooperation effort. [BusinessToday.in]
Commerce and industry minister Piyush Goyal urged India’s tech sector to target a trillion-dollar opportunity by 2035, focusing on AI, data centres, and clean energy infrastructure. [Mint]
India is emerging as a global leader in AI, with a young, tech-savvy workforce and significant contributions to AI talent and projects worldwide. [Ommcom News]
Indus AI Week will be a transformative national event aimed at positioning Pakistan as a global AI contender through nationwide workshops, summits, and initiatives. [ProPakistani]
Luma AI is opening a Riyadh office and partnering with Publicis Groupe Middle East to develop Arabic-native AI for creative communications across MENA. [Business Wire]
Microsoft has partnered with Capgemini to strengthen its sovereign cloud services, helping European organizations maintain data sovereignty and compliance. [Yahoo Finance]
Mistral AI, a French AI startup, is investing €1.2B in Swedish data centres to promote European AI unity and compete globally. [Tech.eu]
Morocco has appointed AI expert Amal El Fallah Seghrouchni as Minister of Digital Transition, aiming to develop AI that serves marginalized populations and languages. [Rest of World]
Omantel launched Otech, a strategic technology subsidiary aimed at transforming Oman’s digital ecosystem and supporting national technological innovation and economic diversification. [The Arabian Stories]
OpenAI is collaborating with G42 to develop a UAE-specific version of ChatGPT tailored to local language, culture, and political sensitivities. [Semafor]
Pakistan Digital Authority and DFINITY partnered to establish a sovereign cloud infrastructure and AI platform, enabling secure, national-scale digital applications in Pakistan. [Business Wire]
Singapore will establish a National AI Council chaired by Prime Minister Wong to drive AI adoption across four key sectors and support workforce transformation. [The Edge]
And Singapore is launching a ‘Champions of AI’ programme to support comprehensive AI transformation across businesses, with expanded tax incentives and grants. [DealStreetAsia]
Suite Numérique is France’s ambitious open-source productivity tool suite, designed to provide sovereign digital workspace alternatives to Microsoft and Google for government services. [WebProNews]
Feature Creeps
Amazon’s new Send to Alexa Plus feature allows Kindle Scribe owners to transform handwritten notes and documents into summaries, tasks, reminders, and brainstorming aids. [The Verge]
Anthropic launched Claude Fast Mode, offering 2.5x faster responses at up to 6x the standard pricing for time-critical tasks. [The Decoder]
And Anthropic has made Claude’s most advanced features, including file creation, Connectors, and Skills, freely available to all users without ads. [TechRadar]
Google is reportedly testing a new Gemini feature that may let users import past conversations from ChatGPT and other AI tools. [eWeek]
Microsoft has launched OneDrive Agents, allowing users to bundle up to 20 files into a single AI-powered .agent file for enhanced document querying and collaboration. [TechRadar]
OpenAI has introduced Server-side Compaction, Hosted Shell Containers, and Skills, enabling AI agents to maintain context and perform complex, long-running tasks with unprecedented stability and flexibility. [VentureBeat]
OpenAI has upgraded Deep Research in ChatGPT to run on GPT-5.2, enabling users to search specific websites and connect apps with enhanced features. [The Decoder]
And OpenAI is enhancing ChatGPT’s deep research tool with a full-screen viewer that enables easier navigation and source tracking for AI-generated reports. [The Verge]
WordPress has launched a new Claude connector, allowing site owners to share back-end data with Anthropic’s chatbot while maintaining control over data access. [TechCrunch]
Zoom has unveiled AI Companion 3.0, offering proactive productivity tools like ‘My Notes’ and ‘Personal Workflows’ to automate tasks across meeting platforms. [TechRadar]
Hype Bubble?
The Magnificent Seven stocks lost nearly US$3T in market value as investors worried about escalating AI spending and rising costs. [Yahoo Finance]
Cloud computing providers have amassed an US$11T backlog, reflecting massive enterprise commitments to AI-driven cloud infrastructure and future digital transformation. [WebProNews]
Research by Cynozure reveals that despite widespread AI adoption, only 15% of organizations can prove its financial impact, highlighting measurement and strategy challenges. [Business Wire]
A Harris Poll survey reveals that 71% of CIOs believe they have until mid-2026 to demonstrate AI’s value or face budget cuts and career risks. [Business Wire]
Microsoft’s Copilot has only 3.3% paid adoption, with 15 million paid seats among 450 million free users, despite CEO Nadella’s claims of widespread use. [TechRadar]
Meanwhile, Microsoft is planning a record $650B AI infrastructure spend in 2026, sparking investor concerns about returns, OpenAI exposure, and Azure capacity. [Yahoo Finance]
But Nvidia CEO Jensen Huang defended massive AI infrastructure spending, arguing that tech companies’ cash flows will rise as demand for computing power surges. [CNBC]
Huang believes the AI infrastructure buildout will continue for seven to eight years, driving massive capital expenditure and transforming computing. [Yahoo Finance]
Goldman Sachs CEO David Solomon downplayed the software industry sell-off, arguing that AI’s impact will create winners and losers, not wholesale disruption. [Yahoo Finance]
Wall Street advanced as tech stocks rebounded, driven by optimism about AI and comments from Nvidia’s CEO. [Yahoo Finance]
Atlassian reassured investors it can add AI to its services without increasing costs, while achieving US$1B in cloud revenue and maintaining customer trust. [The Register]
Oracle shares jumped 10% as analysts pushed back against fears that AI would disrupt the software sector’s business models. [Yahoo Finance]
Big Iron
A Dell’Oro Group report projects worldwide data center capital expenditures will reach US$1.7T by 2030, driven by hyperscalers’ aggressive AI infrastructure investments. [Yahoo Finance]
Alphabet CEO Sundar Pichai warns that meeting surging AI demand is challenging due to constraints in power, land, and supply chain infrastructure. [Yahoo Finance]
AWS is facing significant delays in European data center expansion due to slow grid connection practices, with connection times potentially extending up to seven years. [ITPro]
But Semiconductor Manufacturing International Corp’s Co-CEO warns that tech giants are rapidly building AI data centers without fully understanding their purpose, raising concerns about potential overinvestment. [Gizmodo]
Edinburgh councillors rejected a proposed ‘green’ AI datacenter, citing environmental concerns and questioning the project’s sustainability despite planners’ initial recommendation for approval. [The Register]
Anthropic pledged to cover potential electricity cost increases for consumers as it expands its data center infrastructure to support AI development. [NBC News]
A UC Riverside report reveals California data centers used 50 billion liters of water in 2023, with projections rising to 116 billion liters by 2028. [Mustang News]
Big Tech’s data center expansion is facing increasing legislative resistance due to concerns about power grid strain, electricity costs, and environmental impact. [Yahoo Finance]
New York lawmakers introduced a bill proposing a three-year moratorium on data center development, reflecting a growing bipartisan backlash against such projects nationwide. [Wired]
Australia’s businesses have doubled planned data centre spending to US$52B, positioning the country second only to the US in AI technology investment. [The Sydney Morning Herald]
Firmus secured a $10B debt financing facility from Blackstone and Coatue to expand its AI infrastructure Project Southgate across Australia, aiming to deploy energy-efficient AI Factories. [Verdict]
Google has significantly increased its investment in TeraWulf, supporting the company’s transition from bitcoin mining to low-carbon AI data center infrastructure. [Yahoo Finance]
Meta has expanded its Louisiana AI data center site to nearly 3,700 acres, signaling the company’s massive infrastructure investment in a rural, economically challenged area. [Fortune]
Meta is breaking ground on a US$10B data center in Lebanon, Indiana, as part of its massive AI infrastructure expansion. [Yahoo Finance]
Nebius is planning a massive 240-megawatt data centre in France, with first capacity expected by late summer and half the site operational by 2026. [Yahoo Finance]
Pulsant has invested £10m in a high-density 1.2 MW data center in Milton Keynes, targeting financial, healthcare, IT, and gaming sectors. [ITPro]
Arizton reports New Zealand’s data center market will grow to USD 2.64 billion by 2031, driven by 87% AI adoption and digital transformation. [EIN Presswire]
And China plans to build space-based data centers and digital infrastructure as part of its ambitious space+ initiative within the next few years. [TechRepublic]
Hot Chips
The Semiconductor Industry Association predicted global chip sales will reach US$1T in 2026, driven by robust AI-related demand across the industry. [Yahoo Finance]
Amazon is expanding its partnership with STMicroelectronics to secure chip supply for future data centre infrastructure, including power and analogue components. [Yahoo Finance]
ByteDance is developing its own AI chip with Samsung, aiming to secure advanced computing power and reduce reliance on external semiconductor suppliers. [The Sri Lanka Guardian]
Cisco unveiled the Silicon One G300 chip, designed to accelerate data movement in AI infrastructure and compete for a share of the US$600B market. [Yahoo Finance]
Fractile, a UK chip startup, is expanding its Bristol and London sites with £100m to develop next-generation AI acceleration chips that could challenge GPU-based inference technologies. [Computer Weekly]
Intel is aggressively rebuilding its GPU capabilities by hiring key technical talent and committing to internal development to challenge Nvidia’s market leadership. [TechRadar]
Intel is investing US$100m in SambaNova Systems and hiring Eric Demers to strengthen its AI chip and GPU development strategy. [Yahoo Finance]
Meanwhile, Intel and AMD have warned Chinese customers of server CPU supply shortages, with delivery lead times extending up to six months due to booming AI infrastructure investment. [Yahoo Finance]
Micron Technology plans a US$24B Singapore fab to boost NAND and HBM memory for AI, despite reportedly being excluded from Nvidia’s HBM4 supply chain. [Yahoo Finance]
Micron’s stock dropped after Samsung Electronics secured Nvidia’s certification to produce next-generation HBM4 chips. [Yahoo Finance]
Nvidia is prioritizing AI chip production over consumer GPUs, potentially delaying graphics card releases and making them increasingly unaffordable for gamers. [Gizmodo]
Nvidia’s stock soared as CEO Jensen Huang highlighted surging AI chip demand amid massive cloud provider spending plans. [Investopedia]
Analysis by Nvidia reveals that four inference providers achieved 4x to 10x cost reductions by combining Blackwell hardware, optimized software, and open-source models across various industries. [VentureBeat]
Qualcomm unveiled a 2-nanometre chip in Bengaluru, marking a significant milestone for India’s semiconductor and technology industry. [Mathrubhumi]
Samsung Electronics has started mass production of next-generation HBM4 memory chips designed to power AI data centers at unprecedented speeds. [Yahoo Finance]
Tesla is developing its own AI5 chip and reviving Dojo 3, potentially reducing reliance on Nvidia and sparking a rivalry in autonomous vehicle technology. [Money Digest]
The Trump administration is considering exempting tech giants from chip tariffs in exchange for TSMC’s US investment commitments. [Yahoo Finance]
Warm Bodies
Salesforce has cut nearly 1,000 jobs across multiple teams, including marketing and product management, amid leadership changes and AI expansion efforts. [Firstpost]
Heineken plans to cut up to 7% of its workforce, partly due to AI, following a 2.4% decline in beer volumes in 2025. [CNBC]
Washington State experienced a massive surge in tech layoffs, with over 19,500 job cuts in early 2026 as companies restructure around AI. [Seattle Red]
A survey by Challenger, Gray & Christmas revealed US employers cut over 108,000 jobs in January, the worst start since 2009, with AI potentially contributing to the trend. [Yahoo Finance]
However, when New York state added an AI option to layoff notices, no companies cited technological innovation as the reason for workforce reductions. [Wired]
OthersideAI’s CEO warns that AI could replace up to half of entry-level white-collar jobs within five years, urging professionals to adapt. [Madhyamam]
Khan Academy CEO Salman Khan warns that even a 10% reduction in white-collar jobs due to AI could trigger a depression-like economic crisis. [Yahoo Finance]
But Anthropic cofounder Daniela Amodei argues that humanities and human skills will become more critical in the age of AI, not less. [Yahoo Finance]
And Superhuman CEO Shishir Mehrotra envisions AI as a job expander, not replacer. [Yahoo Finance]
UJET’s CEO argues AI will enhance, not replace, call center agents by streamlining their software tools and improving customer experience. [The Register]
IBM will triple entry-level hiring in the US by 2026, arguing that AI requires reimagining early-career roles and maintaining a robust talent pipeline. [Yahoo Finance]
Meanwhile, A UC Berkeley study reveals that AI tools are causing employees to work more hours, not less, potentially leading to increased burnout and stress. [TechCrunch]
And the UK’s Department for Work and Pensions is exploring chatbot work coaches for Universal Credit claimants while AI threatens to increase job losses across the UK economy. [The Register]
Moving On
Anthropic’s AI safety lead Mrinank Sharma has resigned, citing global crises and a desire to pursue work more aligned with his personal values and sense of integrity. [Yahoo Finance]
OpenAI has hired Dylan Scandinaro from Anthropic to lead AI safety efforts. [The Decoder]
OpenAI fired Ryan Beiermeister, its VP of product policy, after a sexual discrimination allegation related to concerns about a planned ChatGPT ‘adult mode’ feature. [TechCrunch]
And OpenAI researcher Zoë Hitzig resigned, warning that ChatGPT ads could repeat Facebook’s problematic data privacy mistakes. [Ars Technica]
Meanwhile, xAI has lost six of its original 12 co-founders in a week, with Elon Musk framing the exits as a necessary reorganization to improve company effectiveness. [TechCrunch]
Consumer AI
A global memory chip shortage driven by AI infrastructure expansion is causing rising prices for consumer electronics and potentially threatening budget smartphone markets. [Rest of World]
Sam Altman and Jony Ive’s AI hardware venture has encountered delays, trademark issues, and technical challenges, pushing its first device’s release to February 2027. [Gizmodo]
Amazon’s redesigned Alexa app prioritizes its AI chatbot over usability, frustrating users and pushing them to alternative list-management apps like Apple’s Reminders. [The Verge]
Apple is considering allowing third-party AI chatbots like ChatGPT and Gemini to appear on CarPlay, while keeping Siri as the default assistant. [TechRadar]
Apple is exploring AI-powered wearable devices, including a potential pin and smart glasses, after acquiring Q.ai and signalling interest in new AI-enabled product categories. [Gizmodo]
EssilorLuxottica more than tripled its Meta AI glasses sales to 7 million units last year, signaling growing consumer interest in wearable technology. [CNBC]
Even Realities G2 smart glasses offer a larger display, AI integration, and optional R1 smart ring, but software stability remains a persistent challenge for this innovative wearable technology. [Wired]
Sony has introduced premium wireless earbuds with a redesigned look and superior noise cancellation that challenges Apple’s AirPods Pro 3. [Bloomberg]
Starlink denies developing a phone but hints at a future AI-powered device that would radically reimagine mobile technology’s purpose and functionality. [TechRadar]
It’s Only a Model
Alibaba, Baidu, and ByteDance are preparing to unveil advanced AI models during Lunar New Year, intensifying China’s domestic AI competition. [Firstpost]
Alibaba released Zvec, an open-source, in-process vector database library designed for edge and on-device retrieval workloads, offering high-performance vector search with embedded execution. [Marktechpost Media]
Anthropic’s Claude Opus 4.6 tops the Artificial Analysis Intelligence Index, demonstrating strong performance across coding, agent tasks, and scientific reasoning tests. [The Decoder]
ByteDance’s Seedance 2.0 impresses with advanced AI video generation capabilities, showcasing cinematic shots, audio sync, and versatile input handling. [The Rundown AI]
DeepSeek expanded its AI model’s context window from 128,000 to over 1 million tokens, enhancing its ability to process and remember information. [South China Morning Post]
DeepSeek and other Chinese open-source AI models are rapidly gaining global traction, offering high-performance, affordable alternatives to Western proprietary systems while driving innovation through transparency and accessibility. [MIT Technology Review]
Google’s Gemini 3 Deep Think is a specialized reasoning mode that advances scientific research and engineering by solving complex challenges with mathematical rigor and practical applications. [Google]
Meta is developing a highly efficient AI model called Avocado, which outperforms existing models and signals a potential comeback after previous challenges with Llama 4. [The Decoder]
MiniMax launched M2.5, a cost-effective AI language model that promises to transform enterprise tasks by delivering high-performance at a fraction of traditional pricing. [VentureBeat]
OpenAI launched GPT-5.3-Codex-Spark, a low-latency coding model running on Cerebras Systems’ hardware, marking a strategic diversification from its traditional Nvidia infrastructure. [VentureBeat]
Z.ai launched GLM-5, an open-weight AI model with 744B parameters, claiming best-in-class performance and competitive pricing. [Bay Area Times]
Whose Data?
Microsoft launched its Publisher Content Marketplace, a licensing hub for content providers to enable AI training while retaining editorial independence. [Publishers Weekly]
And Amazon is planning to launch a marketplace for publishers to sell content to AI firms. [Yahoo Finance]
Bondu leaked personal data and chat logs of 50,000 toddlers through an unsecured web portal accessible by anyone with a Gmail account. [TechDirt]
Google faces an EU antitrust complaint from publishers over AI Overviews, alleging unfair content use without compensation or consent. [Reuters]
Google has launched WAXAL, a collaborative data set of 21 African languages owned by local partners to improve AI language understanding and digital sovereignty. [Rest of World]
Publishers Hachette and Cengage seek to join a class action copyright infringement lawsuit against Google’s Gemini, arguing they have crucial rights and interests at stake. [Publishers Weekly]
The LLM Ecosystem
Amatrium has launched a 9-language interface and LLM Selector for AmatriumGPT, enhancing global accessibility and enterprise AI control. [EIN Presswire]
Apache HugeGraph became an Apache Top-Level Project, offering a full-stack graph database platform with AI and computing capabilities. [GlobeNewswire]
Automat-it has launched an LLM Selection Optimizer that helps startups cut AI infrastructure costs by up to 60% through data-backed benchmarking. [EIN Presswire]
Backboard.io has become the first AI platform to lead both major memory benchmarks, demonstrating unprecedented performance in long-term AI memory retention and reasoning. [EIN Presswire]
Coforge expanded CodeInsightAI, an AI-powered platform that automates legacy system modernization through advanced code intelligence and engineering capabilities. [Business Wire]
Eccentex launched an AI orchestration platform that embeds intelligent automation into case work, enhancing human decision-making across various industries. [EIN Presswire]
GigaSpaces integrated its eRAG platform with Microsoft 365 Copilot, enabling enterprises to leverage operational data in generative AI applications. [EIN Presswire]
Google Research proposes Natively Adaptive Interfaces (NAI), an AI-driven framework that dynamically adapts software interfaces in real-time to individual user abilities and contexts. [Marktechpost Media]
Google and Microsoft launched WebMCP, a proposed web standard that enables websites to expose structured, callable tools directly to AI agents through a new browser API. [VentureBeat]
IOWN Global Forum and Open Compute Project are collaborating to develop an open, scalable AI computing infrastructure spanning from centralized data centers to edge deployments. [Business Wire]
MinIO has made its AIStor Tables feature generally available, unifying structured and unstructured data in a single object storage platform using Apache Iceberg. [Blocks & Files]
Mastra has developed observational memory, an open-source technology that compresses conversation history using two background agents, enabling stable, cost-effective long-running AI agent workflows. [VentureBeat]
MIT researchers discovered that LLM ranking platforms can be dramatically skewed by removing just a few user interactions, potentially misleading companies seeking the best model for their needs. [MIT News]
Researchers at MIT and ETH Zurich developed a self-distillation fine-tuning technique enabling LLMs to learn new skills without forgetting existing capabilities. [VentureBeat]
Moderne launched Prethink, an AI agent capability that provides structured knowledge to improve reasoning and efficiency in complex software development environments. [Business Wire]
Nighthawk Cyber is developing DATAVARI, an AI monitoring software designed to provide oversight and safety guardrails for AI technologies. [Spectrum News 13]
Nvidia has developed dynamic memory sparsification (DMS), a technique that reduces LLM memory costs by up to eight times while maintaining or improving reasoning capabilities. [VentureBeat]
Portal26 launched an AI Value Realization Solution to help enterprises improve low returns on generative AI investments by analyzing user behavior data. [Business Wire]
SentinelOne expanded its AI Security Platform with new Data Security Posture Management capabilities to secure AI systems from data to runtime. [Business Wire]
Sixwatch launched AI Enablement services to help mid-sized businesses adopt Microsoft 365 Copilot securely and effectively. [EIN Presswire]
Station F has launched an AI accelerator partnering major tech companies to support European startups building applications on foundational AI models. [Wired]
Understand Tech launched AI-In-a-Box, an on-premise enterprise AI appliance enabling secure, offline AI workflows without cloud dependencies. [EIN Presswire]
Veritone is positioning itself as a critical AI agent orchestration platform, offering data governance, unstructured data processing, and workflow management for enterprises. [Business Wire]
Agentic AI
ACTi launched an AI sales platform that provides scalable, 24/7 sales agents to help businesses overcome hiring challenges. [Business Wire]
Alethea expanded its Artemis platform with an AI-powered mitigation agent suite to help organizations quickly respond to digital threats. [PRWeb]
Bellagent launched an AI agent platform designed to help businesses quickly deploy practical, secure AI solutions across various operational functions without complex integrations. [Business Wire]
Genesys launched the industry’s first agentic virtual agent using large action models, enabling autonomous, end-to-end customer request resolution across enterprise systems. [Business Wire]
Goldman Sachs has been collaborating with Anthropic to develop AI agents for automating accounting, trade reconciliation, and client onboarding processes. [CNBC]
Google announced the Agent-to-Agent (A2A) protocol, enabling AI agents to communicate, collaborate, and exchange information securely across different applications and workflows. [The Next Web]
Kentico launched the AIRA Agentic Marketing Suite, introducing specialized AI agents to streamline marketing tasks within its Xperience platform. [EIN Presswire]
LatticeFlow AI and Unique AI have developed a technical blueprint to govern agentic AI in financial services, aligning with FINMA guidance and providing audit-ready evidence. [Business Wire]
Phenom Cloud launched Lexy, an AI-powered suite of digital consultants designed to automate and optimize enterprise People Operations across various HR functions. [EIN Presswire]
Mizo launched an AI-powered service desk platform that autonomously resolves routine IT support tickets across multiple technology platforms. [PRWeb]
Qlik launched agentic analytics in Qlik Cloud, enabling trusted, context-aware AI insights through Qlik Answers and a new MCP server for third-party assistants. [Business Wire]
Zania launched an autonomous third-party risk management platform using AI agents to enable comprehensive vendor assessments without increasing headcount. [Business Wire]
Other LLM Sightings
10turtle has launched AI-powered digital solutions to help small businesses compete effectively by combining AI with creative expertise. [EIN Presswire]
15Five launched Amaya, an AI-powered HR assistant that provides personalized insights and recommendations using company performance data within its platform. [Business Wire]
ACA Group launched Encore AI, an AI-powered tool that enhances marketing review processes by improving compliance accuracy and efficiency. [Business Wire]
Alibaba is partnering with the International Olympic Committee to provide AI-powered LLM assistance for broadcasting and public services at the Milan-Cortina Winter Games. [Caixin.com]
Alliance Creative Group has launched an AI-powered marketing solution for car dealerships, using AI characters to improve lead conversion and engagement. [EIN Presswire]
Altruist launched an AI-powered tax planning feature in Hazel that helps advisors quickly generate personalized tax strategies by analyzing client documents. [Business Wire]
An Adam App’s BABE AI platform by automates construction project planning by converting estimates into comprehensive plans in seconds. [Business Wire]
Augment launched Knowledge Hub, an AI-powered platform that captures and delivers logistics operational insights directly into workflows for brokers, carriers, and shippers. [Business Wire]
Aurasell launched an AI-Native Go-To-Market Operating System that automates workflows across existing CRM platforms without requiring system replacement. [GlobeNewswire]
BridgeView Marketing launched PR Rosetta Stone, an AI-enabled system providing comprehensive, data-driven insights into earned media performance and marketing ROI. [EIN Presswire]
Canopy launched a new bookkeeping intelligence module designed to streamline month-end close processes and provide real-time financial insights for accounting firms. [Business Wire]
CapRelo developed CoreTech, an AI platform that simplifies global relocation services by providing intelligent, personalized support for employees and HR teams. [Business Wire]
Claro Advisors launched an AI-native wealth management platform designed to help independent financial advisors enhance efficiency and client service through integrated technology tools. [Business Wire]
Complexio has deployed its Stevie AI enterprise intelligence layer across five global shipping companies, enabling faster, context-aware operational decision-making. [EIN Presswire]
D&H Technologies launched an AI platform designed to deliver compliant, explainable operational intelligence for regulated industries with measurable ROI. [EIN Presswire]
Datarails launched Spend Control, an AI-powered platform giving CFOs comprehensive visibility and control over vendor contracts and subscriptions. [Business Wire]
FourKites launched Loft, an AI platform that orchestrates enterprise systems by combining internal data with external intelligence through an AI developer agent named Sophie. [Business Wire]
Freehand launched an AI platform that autonomously handles supply chain and finance operations, cutting cycle times by up to 90% for Fortune 500 customers. [GlobeNewswire]
Gappy has launched an AI-powered travel concierge that instantly suggests local activities for tourists in Japan based on their current location and time. [EIN Presswire]
HAiCook is an AI-powered cooking app that generates recipes, plans meals, and simplifies cooking decisions for users across various platforms. [PRWeb]
InfiniMind has developed an AI platform that transforms unused video and audio data into structured, queryable business insights for enterprises. [TechCrunch]
Intuit launched an AI-powered, end-to-end ERP solution for mid-market construction businesses, streamlining workflows and improving profitability across complex project operations. [Business Wire]
iSpot launched SAGE, an AI-powered platform using comprehensive ad data to provide marketing intelligence and insights across TV networks and brands. [PRWeb]
Jump partners with Wealth.com to integrate estate planning data into advisors’ meeting preparation, streamlining client conversations and reducing administrative work. [Business Wire]
Lean Solutions Group launched new AI features for LeanTek AgentEdge and LeanTek Connect, enabling operational workflows with expert human oversight. [Business Wire]
Novi AI has launched an independent website as a standalone AI Creation Studio, integrating multiple AI models to streamline content creation for individuals and teams. [PRWeb]
OpenAI deployed a custom ChatGPT version on GenAI.mil, providing 3 million military personnel with secure AI tools for various operational tasks. [Cryptopolitan]
PPWR Copilot launched a tool to help businesses generate country-specific EU packaging documentation and simplify compliance requirements. [EIN Presswire]
Quetext launched an AI Summarizer tool that generates structured summaries and meeting notes in various formats and lengths for students, educators, and professionals. [EIN Presswire]
RegASK launched an AI-assisted label review tool that delivers comprehensive compliance reports across global markets in seconds. [Business Wire]
ReposiTrak has developed a patent-pending technology that automatically detects and corrects errors in food traceability data with high accuracy. [Business Wire]
Similarweb launched AI Studio, an enterprise AI intelligence solution that provides instant, comprehensive market research insights to employees across an organization. [Business Wire]
Spire.AI launched Knowra, a context intelligence platform that helps enterprises transform fragmented data and AI systems into coherent, decision-ready outcomes. [EIN Presswire]
ToltIQ and Deal Engine partnered to connect deal origination with diligence execution, enabling seamless intelligence sharing across the private equity deal lifecycle. [GlobeNewswire]
Uber launched Cart Assistant, an AI-powered grocery shopping feature in Uber Eats that helps users create lists using text or image prompts. [The Verge]
Veryon has signed an agreement with Saudia Group to provide fleet analytics capabilities using AI-driven defect analysis technology. [Business Wire]
Risks and Responses
The second international AI safety report reveals deep uncertainty about the technology’s trajectory, potential risks, and systemic impacts across jobs, autonomy, and global inequality. [Computer Weekly]
AIM Intelligence successfully breached Anthropic’s Claude Opus 4.6 AI model in 30 minutes, exposing critical security vulnerabilities in autonomous AI systems. [EIN Presswire]
Anthropic is wrestling with the paradox of developing potentially dangerous AI while simultaneously pioneering safety research, hoping its AI model Claude will resolve this tension through ethical self-guidance. [Wired]
Anthropic’s Claude Opus 4.6 system card reveals detailed prompt injection vulnerability rates across different surfaces, with attack success rates ranging from 0% to 78.6% depending on environment and attempts. [VentureBeat]
Anthropic’s Claude Opus 4.6 discovered over 500 high-severity security flaws in open-source libraries by reasoning like a human researcher. [TechRadar]
Sam Altman predicts AI agents will integrate services with or without official APIs, potentially disrupting traditional software business models. [The Decoder]
Astrix Security has released OpenClaw Scanner, a free tool to help organizations detect and mitigate security risks posed by autonomous AI agents like OpenClaw. [Yahoo Finance]
Google Translate, powered by Gemini, is vulnerable to prompt injection attacks that can bypass translation and generate dangerous content. [The Decoder]
Google reports widespread ‘distillation attacks’ on Gemini, with actors attempting to extract its AI model’s inner workings to gain competitive advantage. [NBC News]
Also, OpenAI warned US lawmakers that Chinese rival DeepSeek is using sophisticated methods to extract and copy results from leading US AI models for training. [Yahoo Finance]
Google reports Chinese APT31 hackers used Gemini AI to analyze vulnerabilities and plan cyberattacks against US organizations, highlighting growing AI-powered cyber threats. [The Register]
Korean tech giants Kakao, Naver, and Karrot Market are restricting OpenClaw’s use within corporate networks due to escalating security and data privacy concerns. [The Korea Times]
LayerX discovered a zero-click remote code execution vulnerability in Claude Desktop Extensions that allows malicious Google Calendar events to trigger harmful code execution. [The Register]
Microsoft’s research reveals how AI models can be easily unaligned from safety guardrails through minimal prompting, challenging existing assumptions about AI safety training. [ZDNet]
And a survey by Microsoft reveals that while 80% of Fortune 500 companies are deploying AI agents, only 47% have security controls, creating significant cybersecurity vulnerabilities. [Yahoo Finance]
Microsoft launched a scanner to detect hidden backdoors in LLMs, aiming to identify potential malicious behavior embedded during training. [TechRadar]
Microsoft warns that hidden prompts in AI buttons and links can manipulate AI models to produce biased or misleading recommendations across various critical topics. [The Register]
Midas, backed by OpenAI, Tesla, and SpaceX investors, is developing mathematical verification infrastructure to enhance AI system reliability and security. [Business Wire]
NanoClaw, a lightweight and secure open-source AI assistant developed by Gavriel Cohen, offers container-based isolation and minimalist design to address security concerns in agent frameworks. [VentureBeat]
The Open Data Institute finds LLMs provide unreliable and potentially misleading answers about critical public services and government information. [Computer Weekly]
OpenAI has disbanded its mission alignment team, with its leader Josh Achiam transitioning to a new role as the company’s chief futurist. [TechCrunch]
OpenClaw exposed over 135,000 internet-facing instances, revealing massive security vulnerabilities in the open-source AI agent platform. [The Register]
PromptArmor discovered that AI agents in messaging apps can leak sensitive data through zero-click link previews without user interaction. [The Register]
Regulation
Andreessen Horowitz has become a key influencer of AI policy in Washington during the Trump administration, advocating for minimal regulation. [Bloomberg]
Anthropic has donated US$20m to a super PAC aimed at countering OpenAI’s political influence and promoting AI safety regulation. [The New York Times]
Connecticut lawmakers are pushing to regulate online activity and data privacy for minors, focusing on AI technologies and potential risks to residents. [The Hartford Courant]
Brussels has accused Meta of anti-competitive behavior by blocking rival AI chatbots from WhatsApp, potentially prompting emergency regulatory intervention. [The Register]
The EU is intensifying its tech regulation efforts, challenging US tech giants with hefty fines and new rules, while facing potential trade war threats from the Trump administration. [European Business Magazine]
New York is considering bills to label AI-generated news and pause data center construction due to rising energy demands. [The Verge]
OpenAI is facing allegations of violating California’s AI safety law with its GPT-5.3-Codex model, potentially risking millions in fines. [Fortune]
The UK is introducing legislation requiring frontier AI developers to register, conduct safety tests, and report incidents, marking a shift from voluntary to mandatory AI regulation. [WebProNews]
UK regulators warned that lack of resources and coordination, rather than legal powers, hinder effective AI regulation and protection of human rights. [Computer Weekly]
Conversational AI
ABNB Federal Credit Union launched an AI Voice digital assistant called ‘Digi’ to help members quickly access banking information and reduce wait times. [EIN Presswire]
Ajelix has launched an agentic AI chat platform that autonomously executes business workflows across multiple formats, moving beyond conversational advice to complete task implementation. [EIN Presswire]
Bowtie Solutions and Toma partnered to help automotive dealerships manage service calls using AI and human agents more flexibly and intelligently. [PRWeb]
Bright Pattern and Xceed partnered to deliver AI-powered, omnichannel contact center solutions, enhancing customer experience through advanced technology and operational expertise. [EIN Presswire]
C&M Executive has expanded its use of conversational AI to enhance customer service efficiency and experience through more natural, contextual interactions. [EIN Presswire]
CRMIT Solutions has launched Agent Crucible, an AI-powered testing tool that automates comprehensive, human-like testing of conversational AI agents. [EIN Presswire]
Dassault Systèmes introduced Virtual Companions, AI-powered experts that collaborate with humans to solve complex industrial challenges across disciplines. [Business Wire]
Experian launched an Insurance Marketplace app on ChatGPT, enabling users to compare auto insurance rates through conversational AI. [Business Wire]
FlavorCloud launched XB AI, an intelligent agent designed to simplify global trade complexities for merchants through AI-powered cross-border commerce solutions. [EIN Presswire]
Gradient Labs launched an AI outbound support agent to help financial institutions proactively manage fraud prevention, document collection, payments, and collections. [PRWeb]
Hexnode upgraded its Genie AI to provide IT teams with conversational control, instant device insights, and automated troubleshooting through natural language interactions. [Business Wire]
Intercontinental Exchange has enhanced its MSP mortgage servicing system with a new user experience that introduces advanced automation and productivity tools for servicers. [Business Wire]
Lowe’s is testing AI voice agents in stores to handle customer calls, freeing up staff for more meaningful interactions. [Retail Brew]
Macquarie Bank launched Q, an AI-powered customer support agent offering personalized, 24/7 banking assistance with seamless human hand-off capabilities. [Computer Weekly]
Mobileum and NOHOLD launched a white-labeled AI assistant solution for telecom operators to help SME/SMBs enhance customer engagement and drive sales. [Business Wire]
NiCE’s research reveals enterprises are deploying Agentic AI in customer experience, achieving faster deployments, higher containment rates, and improved satisfaction. [Business Wire]
Nutshell has launched AI-powered Smart Nudges and an Engagement Bar to transform website visitors into leads through automated, context-aware interactions. [PRWeb]
OnceHub launched Phone Booking, an AI-powered phone answering feature that enables customers to schedule meetings through natural conversations. [PRWeb]
OneFii has launched an AI-native operating system with Liv AI, a personalized voice-enabled companion that autonomously manages transportation, commerce, and customer services. [EIN Presswire]
Redfin has launched an app in ChatGPT that allows users to search for homes and explore housing market insights through conversational AI. [Business Wire]
Sabre Corp has partnered with Biztrip AI to develop AI-powered corporate travel assistants using advanced language processing and technology platforms. [Yahoo Finance]
Shopify is well-positioned to benefit from AI-driven conversational commerce, potentially reshaping online retail and attracting merchants through innovative e-commerce platforms. [Yahoo Finance]
SoundHound AI partnered with Bridgepointe Technologies to accelerate enterprise adoption of its AI agent platform and Autonomics solution for streamlined operations. [Yahoo Finance]
Square AI launched a free, conversational AI assistant to help UK small businesses make data-driven decisions more confidently and efficiently. [Business Wire]
VCONIC has partnered with Speechmatics to provide advanced conversation intelligence solutions for healthcare and financial services industries. [GlobeNewswire]
Be Real
Actor Anila Bisha is suing Albania’s government to stop using her image and voice for the AI minister Diella without her consent. [Associated Press]
CoupleRef has launched an AI-powered couples mediation platform that provides live, neutral, and affordable relationship support ahead of Valentine’s Day. [EIN Presswire]
FINQ launched two AI-managed US large-cap equity ETFs, claiming its data-driven model can make better investment decisions than humans. [Yahoo Finance]
Meta was granted a patent for an AI-powered bot that could simulate a user’s social media activity after their death, though the company says it has no plans to implement the technology. [Business Insider]
Nayla Grace, an AI-generated neo-soul artist, has launched two singles with a third coming on February 13, 2026, focusing on emotional depth and intimate musical expression. [EIN Presswire]
Suffescom has launched an AI Companion Platform that offers customizable, adaptable, and compliant conversational AI solutions for businesses seeking personalized digital interactions. [EIN Presswire]
Voice News
aiOla launched QUASAR, an intelligent speech recognition gateway that dynamically routes audio requests to the most suitable ASR engine based on context. [Speech Technology Magazine]
Bartronics is preparing to launch a multilingual, voice-first AI agri-tech app in India in March, enabling farmers to access advisories and support across multiple languages. [Telangana Today]
ElevenLabs has consolidated its AI-powered audiobook services into three tiers, aiming to expand its voice library to 6,000 voices. [Publishers Weekly]
And ElevenLabs has backed Airminal with a grant, supporting its platform that creates AI-voice interfaces from a single prompt for any business and device. [EIN Presswire]
IngenID and Corsound AI partnered to provide advanced voice security solutions, combining biometric authentication and deepfake detection technologies. [PRWeb]
Speechmatics and Boost.ai partnered to enhance enterprise-grade voice AI accuracy and reliability across Europe’s most regulated industries. [GlobeNewswire]
Tavus launched Raven-1, a multimodal AI perception system that enables real-time understanding of human emotion, intent, and context through audio-visual signals. [Business Wire]
Telnyx introduced ClawdTalk, a technical demonstration revealing AI agents’ challenges in real-time voice interactions and execution infrastructure. [GlobeNewswire]
And Telnyx has partnered with Telarus to provide technology distributors access to enterprise-grade voice, connectivity, and AI-powered communication solutions. [GlobeNewswire]
VoiceBox has launched three new AI multimedia services in dubbing, voice cloning, and audio interpretation across 82 languages. [EIN Presswire]
Document AI
Alkymi launched a private credit AI workflow platform to help financial services firms automate document processing, reduce risk, and scale operations more efficiently. [Alkymi]
Automat launched an Intelligent Document Processing Platform that enables teams to extract, classify, and process documents quickly and easily without complex engineering support. [Automat]
Circularo expanded its digital trust platform to support multiple languages, enabling global users to work with secure electronic documents in their native tongues. [EIN Presswire]
Emburse launched Emburse Assurance, an AI-powered expense compliance solution that prevents errors and detects fraud before and after expense submissions. [Business Wire]
Geekflare has expanded from a tech publisher to a software company, launching AI and API tools after reaching 100 million readers. [EIN Presswire]
Google’s PaperBanana uses five specialized AI agents to automatically generate scientific diagrams, outperforming simple image generators in human evaluations. [The Decoder]
Laiye has released ADP SaaS v1.7, offering out-of-the-box document extraction with improved performance, accuracy, and user experience across multiple document types. [Laiye]
Spotify has launched Page Match, a feature using computer vision to sync physical book pages with corresponding audiobook sections. [The Verge]
UpSlide launched an AI Consistency Check feature in PowerPoint to automatically detect errors and reduce review time for financial services firms. [EIN Presswire]
Translation
Acolad has partnered with Gradium to develop enterprise-ready AI interpreting solutions that overcome multilingual communication challenges. [MultiLingual]
ArXiv is requiring English-language submissions, prompting questions about the effectiveness of AI translation tools. [Nature]
OpenAI and Google unveiled new AI translation tools, GPT-Translate and Gemini 3 Pro, with impressive claims but limited transparency about their actual capabilities. [MultiLingual]
T-Mobile is launching a free, network-integrated AI translation feature that enables real-time conversation translation across over 50 languages without requiring additional apps. [Gizmodo]
Wordly launched Workspaces, an AI-powered translation and captioning platform designed to help local governments enhance multilingual communication across operations. [Speech Technology Magazine]
XTM has launched Intelligent Post-Editing, an AI-driven tool that automates translation review while maintaining human oversight and enterprise governance. [MultiLingual]
YouTube has expanded its AI dubbing feature with expressive speech capabilities in eight languages, improving naturalness and giving creators more control. [Slator]
And Nimdzi Insights launched its 2026 language industry survey, expanding to include Language Technology Providers and mid-market players amid AI-driven transformations. [MultiLingual]
Search
Bing Webmaster Tools has launched an AI Performance report that tracks content citations across Microsoft’s AI platforms, providing insights into how and where content is referenced. [Search Engine Land]
Clear Ads has integrated Google’s updated Trends platform with Gemini AI to enhance keyword research for e-commerce clients. [Yahoo Finance]
Google and Bing discouraged creating separate markdown pages for LLMs, warning that such practices could be considered cloaking and violate search engine policies. [Search Engine Land]
Google and Microsoft patents reveal how generative search engines deconstruct queries, evaluate content passages, and build brand understanding through sophisticated AI techniques. [Search Engine Land]
Google has expanded its tool for removing sensitive personal information and nonconsensual images from search results, adding options for IDs and simplifying image removal. [The Verge]
Kind, a new AI desktop platform, allows users to search and generate answers exclusively from their personal, curated data without risking hallucinations or inaccuracies. [GlobeNewswire]
Quire launched AI-powered search and migration tools that transform technical reports into searchable, intelligent resources for environmental and due diligence professionals. [PRWeb]
Reddit says it now has 80 million weekly search users, unifying its search with AI Q&A to become an end-to-end discovery platform for users. [Search Engine Land]
Upland launched AI Conversational Search for Panviva Sidekick, enabling frontline agents to retrieve compliant, accurate knowledge through natural language queries. [Business Wire]
Health Tech
Adit launched an AI Front Desk Agent that handles patient communications, scheduling, and administrative tasks across multiple channels and languages. [EIN Presswire]
Amazon One Medical launched Health Insights, a beta feature that transforms lab results into personalized, actionable health information for eligible members. [Business Wire]
Ambience Healthcare expanded its AI platform’s ‘chart awareness’ to improve clinical documentation, coding, and decision-making across healthcare workflows. [Business Wire]
Apple has reportedly abandoned its AI health coach project, with plans to release individual features through its Health app over time. [Engadget]
Arintra achieved eight-times revenue growth by providing an AI-powered medical coding platform that cuts costs and improves financial performance for healthcare organizations. [Business Wire]
Canary Speech is partnering with Intermountain Ventures to study using vocal biomarkers for early multiple sclerosis detection. [Speech Technology Magazine]
Connect4Patients has launched a healthcare platform with four capabilities aimed at addressing data fragmentation, inefficiencies, and limited patient engagement in the US medical industry. [PRWeb]
DualEnroll.ai launched an AI platform to automate Medicare and Medicaid dual enrollment, reducing costs and processing times by up to 60%. [EIN Presswire]
Fitbit has expanded its Google Gemini-powered AI health coach to iOS users, offering personalized wellness insights and tracking capabilities. [TechRadar]
Hallmark Health Care Solutions has introduced AI agents to automate workforce management tasks and reduce administrative workload for front-line health system leaders. [Business Wire]
i-GENTIC AI has developed AI agents that help MedTech companies improve FDA 510(k) submission success rates by identifying inconsistencies in documentation. [EIN Presswire]
IntelePeer launched SmartAgent Collections, an AI-powered revenue recovery solution that helps healthcare providers collect overdue payments more efficiently and empathetically. [Business Wire]
Martti enabled healthcare interpretation in over 320 languages in 2025, providing rapid, specialized, and culturally sensitive communication for health systems. [EIN Presswire]
Mpathic, a Seattle startup, is expanding its AI safety services by using clinical expertise to reduce dangerous responses in AI models for vulnerable users. [GeekWire]
OttoPilot has launched AI-powered business insights and recommendations to help veterinary clinics improve performance and make data-driven decisions. [Business Wire]
Song & Stark Cooperation has developed SpeechMap, an AI-driven platform that diagnoses childhood language development delays through voice analysis to help parents and rehabilitation centers intervene early. [Dong-A]
Sutter Health has partnered with OpenEvidence to integrate AI-powered, evidence-based medical insights directly into physicians’ electronic health record workflows. [Business Wire]
Vynleads launched the Done With Diabetes app, an AI-powered digital platform providing personalized daily support for type 2 diabetes and metabolic health management. [Business Wire]
WellSky expanded its SkySense AI platform to long-term care and skilled nursing facilities, offering AI-powered tools to improve operational efficiency and clinician productivity. [Business Wire]
A University of Oxford study found AI chatbots provide no better medical advice than search engines, potentially risking patient health by offering inconsistent recommendations. [The Register]
A survey by NBC New York reveals mental health professionals are deeply concerned about AI chatbots’ potential negative impacts on collective mental well-being and individual psychological health. [NBC New York]
Nurses and healthcare professionals are deeply concerned about the potential loss of human connection and emotional support as AI increasingly enters medical settings, fearing technology may diminish critical interpersonal care. [The Indypendent]
Legal Tech
8am expanded LawPay to provide law firms with a unified financial management solution that simplifies billing, payments, and expense tracking. [Business Wire]
DeepJudge has integrated its MCP extension with Claude Cowork, enabling permission-aware search and synthesis across a firm’s prior work product. [Artificial Lawyer]
DISCO has enhanced its Cecilia Q&A tool with scaled agentic AI capabilities for large-scale e-discovery, enabling autonomous, multi-step reasoning across millions of documents. [LawSites]
Docusign has unveiled AI-powered eSignature enhancements to its Intelligent Agreement Management platform, aiming to simplify legal document processing and workflow efficiency. [Yahoo Finance]
Insight Optix launched Legal Hold Evolve, a legal technology solution designed to transform legal hold processes into a strategic advantage for corporate legal teams. [PRWeb]
Legalgain’s whitepaper argues that legal AI requires high-integrity data, domain-specific models, and coordinated reasoning to deliver reliable research. [LawSites]
LegalOn Technologies has launched five AI agents designed to handle a wide range of in-house legal needs, from contract drafting to translation and triage. [Artificial Lawyer]
Romanian rail workers facing corruption charges allegedly sought legal advice from ChatGPT about potential financial damage from their ticket-blocking scheme. [The Register]
Surveill launched a new platform version emphasizing human-led AI for legal and compliance, with five core technological components designed to provide flexible, precise decision-making support. [EIN Presswire]
Thomson Reuters is leveraging proprietary legal data and AI-enabled tools like Westlaw Advantage to defend against potential AI disruption in the legal market. [Yahoo Finance]
AI legal advice is driving lawyers bananas. [Axios]
Ed Tech
Doris, a UK-based EdTech platform for international school discovery, is expanding globally after successful launch in Southeast Asia. [The Daily Express]
Indian PM Modi emphasized balanced AI use in education, warning students against overdependence while highlighting technology’s supportive role in learning. [News9 LIVE]
Israel Sci-Tech Schools introduced ‘Buddy’, an AI-powered teaching-support robot that assists teachers and helps students in Kiryat Bialik classrooms. [SRN News]
Optivara launched an Insights Platform that helps higher education institutions understand and improve their visibility in AI-generated answers about college search and admissions. [PRWeb]
Tranquility AI has partnered with Mount Aloysius College to introduce AI-powered investigative tools into criminology education through its TimePilot platform. [Business Wire]
Udemy has partnered with Glean to integrate AI-driven, context-aware learning directly into workplace systems, helping professionals upskill more effectively. [Business Wire]
And Udemy partnered with OpenAI to integrate its comprehensive skills training content directly into ChatGPT, enabling personalized and interactive learning experiences. [Business Wire]
Funding
AdZen secured a funding round to develop privacy-friendly, content-focused advertising solutions for conversational AI platforms and web publishers. [EIN Presswire]
Alphabet launched a US$15B multi-tranche bond offering, including a potential 100-year bond, to support its massive AI-driven capital expenditure plans. [Yahoo Finance]
Anthropic is seeking to raise US$20B in new capital at a US$350B valuation, with strong investor interest and support from strategic partners like Nvidia and Microsoft. [TechCrunch]
MGX is reportedly in talks to invest hundreds of millions in Anthropic’s funding round. [Tech Funding News]
Blackstone is investing US$200m in Anthropic, increasing its stake to roughly US$1 billion at the AI startup’s current US$350 billion valuation. [Yahoo Finance]
Appficiency received a minority equity investment from Canadian Business Growth Fund to accelerate its global AI-powered enterprise digital transformation services. [Yahoo Finance]
Berget AI raised €2.1m to develop a sovereign AI platform enabling European organizations to build AI applications with local data control. [Tech.eu]
Bracket, a London-based FX and treasury platform, has raised US$7m in seed funding to automate mid-market financial operations through its AI-enabled technology. [Tech.eu]
Bretton AI raised US$75m in Series B funding, rebranding from Greenlite AI to build an AI standard for financial crime operations at regulated institutions. [Business Wire]
Cerebras Systems raised US$1B at a US$23B valuation, with Benchmark Capital investing heavily in the AI chipmaker’s latest funding round. [TechCrunch]
Complyance launched an AI-powered data compliance platform that automates governance and risk management tasks for enterprises, raising US$20m in Series A funding. [TechCrunch]
Databricks has raised US$5B at a US$134B valuation, bolstering its balance sheet to invest in enterprise AI products. [Yahoo Finance]
Daytona raised US$24m to build infrastructure for large-scale, flexible computing environments designed to support software agents’ complex workload needs. [Tech.eu]
Didero has raised US$30m to deploy AI agents that autonomously execute procurement tasks for manufacturers and distributors. [Yahoo Finance]
Dono raised $6.5M to build AI-powered infrastructure that transforms fragmented US property records into accessible, instant ownership data. [Business Wire]
Electric Twin, an AI platform creating synthetic audience models, has raised US$14m in funding led by Atomico to help organizations better understand and predict human behavior. [Tech.eu]
ElevenLabs raised US$500m in Series D funding at an US$11 billion valuation, led by Sequoia Capital, to expand its AI audio and voice platform globally. [FinSMEs]
Entire, led by former GitHub CEO Thomas Dohmke, raised US$60m to help developers manage AI-generated code through an innovative open-source tool. [TechCrunch]
Firmus Technologies secured US$100m from Maas Group Holdings, advancing its AI data centre expansion across Australia with a focus on sovereign manufacturing. [Startup Daily]
Fractile will invest £100m in expanding its UK operations, creating a new engineering facility in Bristol to develop next-generation AI chip systems. [BusinessLive]
GenFlux raised US$4.2m in seed funding to help brands optimize their visibility in AI-generated search answers. [FinSMEs]
Glean has raised US$150m to develop an AI work assistant that connects internal systems and delivers intelligence across enterprise workflows. [TechCrunch]
Goodfire raised US$150m to develop an AI model design environment that helps researchers understand and control AI systems’ internal mechanisms and behavior. [Tech Funding News]
Harvey is reportedly in talks to raise US$200m at an US$11 billion valuation, continuing its rapid growth in legal AI. [TechCrunch]
Intel is reportedly preparing to invest at least US$100m in SambaNova Systems to strengthen its AI capabilities. [Yahoo Finance]
MARC, an AI real estate asset management startup, has raised US$1m from angel investors to develop contract management technology for large property portfolios. [Tech.eu]
matia raised US$21m in Series A funding to develop a unified data operations platform that simplifies data infrastructure for AI-driven companies. [Business Wire]
Meridian has raised US$17m to develop an AI-powered financial modeling workspace that aims to make spreadsheet analysis more predictable and auditable. [TechCrunch]
Modal Labs is in talks with General Catalyst for a potential US$2.5B funding round, more than doubling its valuation from five months ago. [TechCrunch]
Monaco, a new AI sales platform, has raised US$35m to provide AI-powered sales tools with human expert oversight for early-stage startups. [TechCrunch]
Newo raised US$25m in Series A funding to accelerate its AI platform for turning customer conversations into revenue for service providers. [FinSMEs]
Overmind, a London-based startup developing a supervision layer for AI agents, has raised £2m in seed funding led by Osney Capital. [Tech.eu]
Positron AI, a maker of energy-efficient AI inference hardware, raised US$230m in Series B funding at over US$1B valuation. [FinSMEs]
Reflow has raised over US$15m in seed funding to launch an AI-powered workflow intelligence platform that provides enterprises with real-time visibility into operational processes. [Tech.eu]
Runway has raised a US$315m Series E round, nearly doubling its valuation to US$5.3 billion while focusing on world models and expanding its AI video-generation capabilities. [TechCrunch]
Vista Equity Partners is leading a US$350m funding round in AI chip startup SambaNova Systems, marking a rare departure from its software investment focus. [Reuters]
Scala, an operational intelligence platform for contact centers, has emerged from stealth with US$8.5m in funding to help organizations manage hybrid human-AI customer service environments. [Business Wire]
Seamflow, an AI software company for testing and certification, has raised US$4.5m in seed funding to streamline complex regulatory processes. [Tech.eu]
Andreessen Horowitz has invested in Shizuku AI, a California startup founded by a Japanese entrepreneur developing an AI companion character. [Nikkei Asia]
Simple AI has raised US$14m to develop its voice AI agent platform for automating inbound phone calls with personalized customer interactions. [Pulse 2.0]
Smart Bricks, an AI-powered proptech startup, raised US$5m to help real estate investors analyze deals and complete transactions using advanced AI tools. [TechCrunch]
Tactful AI has raised US$1m in pre-Series A funding to expand its agentic customer experience platform in Egypt and beyond. [FinSMEs]
Vesiro, a Chalmers University spinout, raised €1.6m to develop an Elasticsearch plug-in that improves data search efficiency and reduces server energy consumption. [Tech.eu]
WINN.AI raised US$18m in Series A funding to provide real-time AI guidance for sales teams, helping improve win rates and operational efficiency. [Business Wire]
Acquisitions
BriefCatch has acquired WordRake, integrating its core product and technology assets to expand its legal editing software offerings. [LawSites]
Dcycle has acquired ESG-X, an AI-enabled sustainability reporting software company, to support its European expansion and integrated data infrastructure strategy. [Tech.eu]
Model ML acquired Captide to enhance its AI workflow platform with structured, citable corporate disclosure data for financial institutions. [Business Wire]
Nebius has agreed to acquire Tavily, adding agentic search capabilities to its AI cloud platform to support enterprise AI agent development. [Business Wire]
Phenom acquired Be Applied, an AI-driven cognitive assessment solution that enables skills-first hiring at enterprise scale with fairness and precision. [Business Wire]
Submer acquired Radian Arc to provide full-stack AI infrastructure spanning core datacenters and edge compute for telco operators and enterprises. [Business Wire]
Thomson Reuters has acquired Noetica, an AI-native start-up that transforms transaction-deal data into structured market intelligence for deal professionals. [Legal IT Insider]
UiPath has acquired WorkFusion, enhancing its AI-powered solutions for financial crime compliance and banking workflows. [Business Wire]
There’s More
Anthropic’s Claude Opus 4.6 spent $20,000 and nearly 2,000 sessions creating a partially functional Rust-based C compiler capable of building the Linux kernel. [The Register]
Crypto.com’s founder Kris Marszalek has purchased AI.com for US$70m and launched an AI agent platform aimed at creating a decentralized network of self-improving AI agents. [Gizmodo]
Google DeepMind CEO Demis Hassabis envisions a transformative AI-driven future of scientific discovery, personalized medicine, and potential galactic exploration within the next 15 years. [Yahoo Finance]
Nvidia CEO Jensen Huang claimed in a CNBC interview that generative AI no longer hallucinates. [The Decoder]
QuitGPT launched a campaign urging users to cancel ChatGPT subscriptions over OpenAI’s ties to Trump’s campaign and ICE’s use of its technology. [MIT Technology Review]
The Back Cover
Here’s Elle and Toni with a cover of Steely Dan’s Dirty Work.
If you know of a great cover version of a back-catalogue classic, drop an email to us at backcover@language-technology.com and we’ll consider it for inclusion here.
Got this from someone else?
Did you find this newsletter useful? if so, you might forward it to a friend; or you could email us at news@language-technology.com to tell us what you want more of.
Did you hate this newsletter? if so, you could forward it to an enemy; or you could email us at news@language-technology.com to tell us why – make it better!