This Week in NLP #358
Keep up with what happened in NLP in the week ending Friday 26th September 2025.
Above the Fold
Overwhelmed? Here’s our pick for five things to know about this week.
Anthropic launched a multimillion-dollar ‘Keep thinking’ brand campaign to position Claude as a responsible AI problem-solving partner for consumers. [Axios]
Microsoft has integrated Anthropic‘s Claude AI into Microsoft 365 Copilot, expanding its multi-model approach beyond OpenAI‘s offerings. [The Register]
Nvidia is investing up to US$100B in OpenAI, providing 10 gigawatts of computing power to advance next-generation AI infrastructure. [TechRadar]
Trump signed an executive order imposing a $100,000 fee on new H-1B visa applications, causing concern among Bay Area tech companies and workers. [ABC7 News]
European lawmakers and global leaders have called for binding international rules to prevent dangerous AI applications by establishing global ‘red lines’ by 2026. [Euronews]
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
Alibaba’s stock surged after announcing an expanded AI investment strategy exceeding US$50B. [Bloomberg]
Alibaba Cloud expanded its datacenter presence in Brazil, France, and The Netherlands, aiming to bring its AI capabilities to new global markets. [The Register]
Databricks signed a US$100m deal with OpenAI to integrate its models, including GPT-5, into its data platform and AI product Agent Bricks. [TechCrunch]
Google Cloud has added AI coding startups Lovable and Windsurf to its customer roster, highlighting its growing prominence in the cloud computing market. [TechCrunch]
Google Cloud is strategically courting AI startups with infrastructure, credits, and support, positioning itself to capture the next generation of innovative companies in the rapidly evolving AI landscape. [TechCrunch]
Google expanded its AI Mode search feature to Spanish-speaking users, enabling natural language queries and conversational interactions. [TechCrunch]
And Google has expanded its AI Plus plan to over 40 countries, offering Gemini Pro and additional AI tools at a discounted rate. [TechCrunch]
Meta‘s Llama AI models have been approved by the US General Services Administration for broader use across federal agencies. [ZDNet]
Microsoft is haunted by the prospect of missing the AI era, fearing irrelevance and driving significant cultural and strategic changes within the company. [The Verge]
Microsoft launched Windows AI Labs, a pilot program for rapidly testing and validating new AI features in Windows 11. [TechRadar]
Nvidia‘s market value surged US$150B after announcing a strategic partnership with Intel, generating a 30x return on its US$5B investment. [TechRadar]
OpenAI expanded its budget-friendly ChatGPT Go subscription to Indonesia, offering enhanced features at a competitive price point. [TechCrunch]
OpenAI has developed GDPval, a benchmark testing AI models’ performance against human professionals across nine key industries, revealing promising results for GPT-5 and Claude Opus 4.1. [TechCrunch]
And OpenAI briefly revealed experimental agent models in ChatGPT, suggesting ongoing development of advanced, tool-augmented AI assistants with potential improvements in task completion. [TestingCatalog]
xAI has faced a series of leaks revealing internal turmoil, controversial AI development practices, and staff departures. [Gizmodo]
xAI is suing OpenAI, alleging the company induced former employees to steal trade secrets related to its Grok AI model and rapid data center deployment. [Sherwood]
xAI has secured a contract with the US government to provide its controversial Grok AI to federal agencies for $0.42, despite concerns about its ideological bias and inappropriate content. [The Register]
And xAI‘s Grok chatbot has struggled to gain traction, reaching only 64 million monthly users while facing challenges with controversial responses and leadership changes. [The Decoder]
AI Supremacy
China is aggressively developing AI infrastructure by creating data centre clusters, linking remote processors, and reducing dependence on Western technology through massive investments and homegrown solutions. [Asia Financial]
Huawei unveiled a comprehensive AI infrastructure with SuperPoD technology, challenging Nvidia‘s dominance and advancing China’s technological self-reliance amid US sanctions. [WebProNews]
Nvidia and OpenAI executives supported Trump’s H-1B visa fee increase, emphasizing the importance of attracting global talent to the United States. [CNBC]
The Trump administration, led by David Sacks, is opposing a defense bill provision that would restrict US microchip exports to prioritize domestic customers. [Politico]
US leaders disagree on how to handle AI competition with China, particularly regarding chip sales and regulatory approaches. [Axios]
Sovereign AI
Abu Dhabi has unveiled a comprehensive plan to transform its government into a fully AI-native system by 2027, investing US$3.5B to digitize services and boost economic growth. [The Decoder]
Argentina is losing its top AI talent to international research centers due to limited opportunities, funding challenges, and inadequate infrastructure at home. [Rest of World]
Atlas Power and Sovereign Digital Infrastructure signed a memorandum of understanding to develop sustainable, AI-ready power and data center infrastructure in Canada. [EIN Presswire]
AWS and SAP expanded their collaboration to make SAP Sovereign Cloud capabilities available on AWS’s new European Sovereign Cloud. [Business Wire]
Ghana’s technology minister positioned Africa as a potential AI ‘gold mine’, highlighting the continent’s youthful population and untapped data resources as key to global technological development. [News Ghana]
India is poised to become a unique AI powerhouse, leveraging its large market, developer talent, and innovative approach to create affordable, user-centric AI services. [The Economist]
And India is pursuing an ambitious semiconductor strategy, aiming to develop a comprehensive chip industry through significant investments and policy support, despite facing substantial technological and infrastructural challenges. [CNBC]
SAP has partnered with OpenAI and AWS to develop sovereign cloud services for European governments, focusing on local data storage and compliance. [Bloomberg]
Technology Ireland urged government investment in AI skills training and R&D tax credits to ensure Ireland’s competitive transition to an AI-enabled economy. [Irish Examiner]
The UAE, Saudi Arabia, India, and African nations are reshaping global technology by developing innovative, culturally-attuned AI solutions that challenge traditional tech innovation centers. [Observer]
The UK-US Technology Prosperity Deal attracted £150B in investments from major tech companies, potentially positioning the UK as an AI hub while raising questions about long-term economic benefits. [The Conversation]
The UK affirmed its ability to independently regulate AI despite signing a technology partnership with the US during Trump’s state visit. [Politico]
FTSE 250 executives strongly believe urgent UK grid upgrades are essential to support AI datacentre expansion and maintain global competitiveness. [Computer Weekly]
The US government’s $100,000 H-1B visa fee could trigger a reverse brain drain, potentially boosting India’s startup ecosystem with returning tech talent. [The Economic Times]
Trump’s $100,000 H-1B visa fee creates an opportunity for countries like South Africa to attract global tech talent by offering more welcoming immigration policies. [TechCentral]
Germany, Canada, the UK, and China are actively offering alternative immigration pathways to attract skilled Indian tech talent amid tightening US visa policies. [The Economic Times]
Wadhwani AI Global launched to develop context-specific AI solutions for healthcare, education, and agriculture in developing countries across Africa and Latin America. [Business Wire]
Feature Creeps
Google’s Gemini is launching in Chrome for Pro and Ultra users in the US, transforming web browsing with AI-powered assistance across tabs and Google apps. [TechRadar]
Google updated Gemini 2.5 Flash with improved reasoning, formatting, and image understanding capabilities across its app and developer platforms. [9to5Google]
Google has expanded Gemini CLI and Code Assist access for Pro and Ultra subscribers, offering enhanced development workflow tools. [Google]
Google introduced an AI-powered photo editing feature in Google Photos, allowing Android users to edit images using natural language commands. [TechCrunch]
Google launched Mixboard, an AI-powered mood board creation tool that enables users to generate and edit images using text prompts. [TechCrunch]
And Google’s Ask Gemini in Meet answers questions, summarizes discussions, and helps participants catch up using meeting captions and other resources. [TechRadar]
Plus, Gemini is now integrated into Google Chat, enabling users to refine messages, correct grammar, and improve communication across various professional contexts. [ZDNet]
Google‘s AI Search Live transforms searching into a conversational, camera-enabled experience that could fundamentally change how users interact with information. [TechRadar]
Grammarly has expanded its AI writing assistance to support grammar, rewrites, and translations in five additional languages beyond English. [MultiLingual]
Meta is enhancing Facebook Dating with an AI-powered Dating Assistant and a weekly ‘Meet Cute’ surprise match feature to improve user experience. [TechRadar]
Microsoft added an AI model selector to Visual Studio Code, automatically choosing Claude Sonnet 4 over GPT-5 for optimal coding performance. [The Verge]
And Microsoft added another Copilot button to Windows 11’s taskbar, allowing users to share window contents with Copilot Vision directly. [The Verge]
Notion has released version 3.0, introducing AI agents that can create documents, build databases, and execute multi-step workflows with customizable instructions. [The Decoder]
OpenAI has launched Pulse, a new ChatGPT feature generating personalized morning reports for Pro users, aiming to transform the AI tool into a proactive assistant. [TechCrunch]
Perplexity introduced an AI email assistant for Outlook and Gmail that manages inboxes, drafts replies, and schedules meetings for $200 monthly. [Windows Central]
Hype Bubble?
Bain & Company reports that generative AI in software development has yielded modest productivity gains, requiring a comprehensive rethinking of the development lifecycle to realize its potential. [The Register]
And consultants at Bain warned that the AI infrastructure sector may fall US$800B short of the US$2 trillion investment needed by 2030 to sustain current growth projections. [The Register]
A study by BearingPoint reveals a widening AI adoption gap, with only 8% of organizations successfully scaling AI initiatives and 7% embedding agentic AI enterprise-wide. [Business Wire]
DeepSeek‘s R1 model training cost was reported as US$294,000, but actually required 2.79m GPU hours and approximately US$5.87m. [The Register]
Deutsche Bank warned that AI spending is unsustainably high, potentially propping up the US economy through massive investments with uncertain returns. [Gizmodo]
A survey by Google revealed that 90% of developers now use AI at work, with two-thirds relying heavily on it despite limited trust in its output. [TechRadar]
A study by LoopMe revealed AI adoption has reached approximately 30% in the UK, US, and Australia, with younger generations leading usage. [Business Wire]
Meta‘s Mark Zuckerberg acknowledged potential AI investment bubble risks but believes missing out on transformative technology would be more costly. [Fortune]
Microsoft struggles to convince enterprise customers of the value of Copilot+ PCs, facing tepid sales and scepticism about the necessity of on-device AI technology. [The Register]
A forecast by Omdia shows the agentic AI market rapidly expanding from US$1.5B in 2025 to US$41.8B by 2030, outpacing traditional generative AI growth. [Business Wire]
OpenAI’s Sam Altman suggests AI exists in three bubbles—speculative, infrastructure, and hype—but systematic implementation can still extract significant value from the technology. [Fast Company]
And Altman predicted that by 2030, AI will likely achieve superintelligence capable of making scientific discoveries beyond human abilities. [Politico]
A survey by PagerDuty revealed that 75% of companies have deployed multiple AI agents, with 81% of executives trusting AI to handle critical operational tasks. [Business Wire]
A global survey by Riverbed revealed that despite significant AI investments, only 12% of AI projects have been fully deployed, highlighting implementation challenges. [Business Wire]
Solvd‘s research revealed that 71% of CIOs and CTOs believe leadership has unrealistic expectations about AI’s return on investment. [Business Wire]
Tech giants face potential multi-trillion-dollar valuation adjustments due to rapidly evolving AI infrastructure and uncertain depreciation schedules for expensive semiconductor investments. [The Economist]
Big Iron
CoreWeave has expanded its data center capacity agreement with OpenAI to up to US$6.5B, bringing their total contract value to approximately US$22.4B. [Business Wire]
G42 and Microsoft‘s planned US$1B geothermal-powered data center in Kenya’s Olkaria region has stalled due to unclear business rationale and shifting geopolitical priorities. [Semafor]
Khazna has rapidly become the UAE’s leading data center provider, building critical AI infrastructure that supports global tech giants and advances the country’s digital transformation. [Rest of World]
Lambda has deployed the industry’s first hydrogen-powered Nvidia GB300 NVL72 systems at ECL’s zero-emissions Mountain View data center, marking a sustainable AI infrastructure milestone. [Business Wire]
Meta has expanded its Alabama data center investment to US$1.5B. [Montgomery Advertiser]
Microsoft is building a massive AI datacenter in Wisconsin, investing US$7B to create jobs, support local communities, and advance sustainable technology innovation. [Microsoft]
Nvidia‘s CEO predicts US$3-4 trillion will be spent on AI infrastructure by 2030, with companies like Microsoft, OpenAI, and Oracle investing billions in massive data center projects. [TechCrunch]
OpenAI has partnered with Nvidia to deploy 10 gigawatts of AI infrastructure, with Nvidia potentially investing up to US$100B. [OpenAI]
But Nvidia‘s partnership with OpenAI could potentially create an unfair market advantage and raise significant antitrust concerns. [DealStreetAsia]
And OpenAI announced an expanded Stargate project with Oracle and SoftBank, adding five new data centers to support massive AI infrastructure development across the US [Gizmodo]
Plus, OpenAI aims to build a gigawatt-per-week AI infrastructure factory, believing increased computational power could revolutionize global problem-solving and potentially become a fundamental human right. [TechRadar]
Oracle is reportedly in advanced talks with Meta for a US$20B cloud computing deal to support Meta’s AI infrastructure expansion. [Tech Funding News]
Oracle raised US$18B in debt to fund AI infrastructure expansion, despite concerns about customer funding and potential financial strain. [The Register]
And Oracle faced a warning from Moody’s about significant counterparty risks in its US$455B AI infrastructure buildout, primarily due to potential revenue concentration from a few key clients. [The Register]
The Chips are Up
Alibaba secured China Unicom as a client for its T-Head AI chips, signaling growing momentum in its semiconductor development efforts. [Bloomberg]
FuriosaAI has unveiled an AI server that matches Nvidia‘s H100 performance while consuming significantly less power, delivering 4 petaFLOPS at just 3kW. [TechRadar]
Huawei has unveiled a three-year roadmap to challenge Nvidia‘s AI chip dominance by developing massive, interconnected processor clusters. [TechRepublic]
And Huawei has collaborated with Zhejiang University to modify DeepSeek‘s AI model using Ascend chips, creating a version more compliant with Chinese content regulations. [The Register]
Intel has approached Apple about a potential investment and closer collaboration. [TechCentral]
Nvidia and Intel’s US$5B partnership is aimed at competing with AMD by developing integrated CPU-GPU chips for the notebook market. [The Verge]
Qualcomm launched the Snapdragon 8 Elite Gen 5 chipset, enabling unprecedented 220 tokens-per-second AI processing on mobile devices. [ZDNet]
TSMC has unveiled AI-designed chips that promise to reduce energy consumption by using advanced software and multi-chiplet packaging techniques. [South China Morning Post]
Warm Bodies
Meta has poached Yang Song, a high-ranking OpenAI researcher, to join its Superintelligence Labs amid an ongoing AI talent recruitment effort. [Wired]
OpenAI and Jony Ive’s io startup recruited over two dozen Apple hardware employees to develop potential wearable devices using Apple’s supply chain partners. [9to5Mac]
Trump has proposed a $100,000 fee for new H-1B visas, potentially disrupting technology and finance sectors’ workforce. [CNBC]
xAI has experienced executive departures due to management concerns, conflicts with key advisers, and questions about financial projections and company operations. [The Wall Street Journal]
Consumer AI
Google has completed migrating Nest devices to the Google Home app, signalling a shift in its smart home strategy amid ongoing hardware challenges and anticipation of AI-powered updates. [The Verge]
Google is bringing Gemini to Google TV, enabling conversational AI interactions on smart TVs with context-aware, educational, and entertainment-focused capabilities. [TechRadar]
Meta Ray-Ban Gen 2 glasses face delayed UAE release due to strict photography and cybersecurity laws governing individual privacy and national security concerns. [ITP.NET]
Meta‘s CTO Andrew Bosworth explained the technical reasons behind the live demo failures of the company’s new smart glasses at Meta Connect, attributing them to resource management and server routing issues. [TechCrunch]
OpenAI is exploring development of AI devices including a smart speaker, glasses, voice recorder, and pin, potentially leveraging Apple‘s supply chain and talent. [The Verge]
OpenAI has reportedly partnered with Luxshare to develop a context-aware, pocket-sized AI device challenging traditional smartphone dominance. [The Times of India]
It’s Only a Model
Alibaba launched Qwen3-Omni, an open-source, multimodal AI model that accepts text, image, audio, and video inputs while outputting text and audio across 119 languages. [VentureBeat]
And Alibaba‘s Qwen team released FP8-quantized Qwen3-Next-80B-A3B models with two variants, offering high-throughput inference and ultra-long context capabilities. [Marktechpost Media]
AWS added fully managed Qwen3 and DeepSeek-V3.1 open-weight AI models to Amazon Bedrock. [siliconANGLE]
DeepSeek has released DeepSeek-V3.1-Terminus, an improved LLM with enhanced agent performance, language consistency, and two operational modes. [VentureBeat]
FICO has developed two domain-specific AI foundation models, FICO FLM and FICO FSM, with a unique Trust Score to ensure accuracy and compliance in financial services. [VentureBeat]
Huawei developed DeepSeek-R1-Safe, a government-approved AI model with enhanced security features, created in collaboration with Zhejiang University. [TechRepublic]
Intuit has developed a Generative AI Operating System with custom financial LLMs that improve accuracy, efficiency, and personalization across its platforms. [VentureBeat]
Liquid has developed small, task-specific AI models (Nanos) that deliver high-performance capabilities comparable to much larger models while enabling efficient, low-cost, and privacy-preserving on-device deployment. [GlobeNewswire]
Luma AI unveiled Ray3, a video generation model offering HDR capabilities, advanced reasoning, and visual control for professional-grade video creation. [The Decoder]
Meta FAIR has developed Code World Model, a 32B-parameter LLM trained on execution traces and agent-environment interactions to enhance code generation through world modeling. [Marktechpost Media]
xAI has released Grok 4 Fast, a highly efficient AI model offering near-frontier performance at dramatically lower costs, with a massive 2 million token context window and advanced reasoning capabilities. [VentureBeat]
Whose Data?
Anthropic has filed revised materials addressing Judge Alsup’s concerns about a US$1.5B copyright infringement settlement with authors, proposing a detailed payout and outreach strategy. [Publishers Weekly]
Cloudflare has expanded Project Galileo to provide free AI crawl control and bot management tools to help nonprofits and independent media protect their online content. [Business Wire]
LinkedIn is set to train its AI models on member profiles by default. [TechRadar]
Microsoft has initiated talks with US publishers to create a two-sided marketplace that would compensate them for AI content usage. [Axios]
Neon Mobile rapidly rose in App Store rankings by offering users payment for phone call recordings, which the company then sells to AI companies for ML purposes … [TechCrunch]
… but the app was taken offline after a security flaw exposed users’ phone numbers, call recordings, and transcripts. [TechCrunch]
The LLM Ecosystem
Apple introduced its Foundation Models framework, enabling developers to integrate local AI models into iOS 26 apps, enhancing user experiences with small-scale, privacy-focused features. [TechCrunch]
Clarifai has unveiled a new reasoning engine that optimizes AI model inference, promising faster performance and reduced costs through advanced software techniques. [TechCrunch]
Comet and Amazon SageMaker AI have collaborated to provide enterprises with a comprehensive ML experiment management platform that enables reproducible, scalable, and compliant ML workflows. [Amazon]
Google has launched the Data Commons Model Context Protocol Server, enabling AI systems to access and utilize public datasets through natural language prompts. [TechCrunch]
And Google has released Chrome DevTools MCP, an open protocol server enabling AI coding agents to control and inspect Chrome instances with detailed runtime debugging capabilities. [Marktechpost Media]
GitHub has launched the MCP Registry, a centralized platform for discovering and exploring MCP servers, aiming to simplify AI tool integration and foster an open, interoperable ecosystem. [GitHub]
Microsoft has launched AI agents in GitHub Copilot to automatically modernize legacy Java and .NET applications, dramatically reducing technical debt and enabling enterprise AI transformation. [VentureBeat]
ProCogia launched Data Science in a Box, a secure, multi-cloud solution enabling enterprises to adopt generative AI quickly while maintaining governance and compliance. [PRWeb]
Scale AI introduced a global, crowdsourced AI model leaderboard system that ranks performance across diverse demographics through user voting. [Bloomberg]
Sentry launched an AI code review feature that helps developers predict and prevent bugs before they reach production, leveraging AI to enhance software development. [Business Wire]
Snowflake launched a comprehensive startup program providing AI infrastructure, venture capital partnerships, and resources to help founders build and scale enterprise-grade AI applications. [Business Wire]
ThoughtSpot joined Snowflake and other leaders in launching the Open Semantic Interchange, an open-source initiative to standardize data definitions across platforms. [GlobeNewswire]
Thunk.AI released a ‘Hi-Fi’ benchmark to measure AI automation reliability. [Business Wire]
Trismik has developed an AI evaluation platform using psychometrics and adaptive testing, reducing evaluation costs by up to 95% while providing more nuanced insights into model capabilities. [Tech Funding News]
TwelveLabs launched its MCP Server, enabling AI assistants to understand, search, and interact with video data through a universal adapter. [PRWeb]
Agentic AI
Accelirate has launched AcceliQA, an AI-powered QA testing program promising 40% ROI and 3X faster execution through agent-led automation. [EIN Presswire]
Airia has launched Agent Constraints, a pioneering policy engine enabling centralized, context-aware governance of AI agents without code modifications. [Business Wire]
Algolia launched Agent Studio, a platform enabling developers to build scalable, context-aware AI agents with trusted retrieval, tool orchestration, and observability. [Business Wire]
Amazon has introduced a Nova Act extension for building and testing AI agents directly within integrated development environments. [GeekWire]
Asana has introduced AI Teammates, collaborative agents that understand organizational context and can support multiple teams while adapting from human feedback. [Business Wire]
Coral Protocol has released v1 of its agent stack, providing a standardized framework for discovering, composing, and operating AI agents across different platforms. [Marktechpost Media]
Fotor launched an AI Agent that enables users to create and edit images and videos through intuitive voice and text commands across diverse creative tasks. [EIN Presswire]
Hostinger launched Web2Agent, a free tool that enables websites to become AI-agent-compatible with a single click while preserving design and privacy. [TechRadar]
Klaviyo launched Marketing Agent and Customer Agent, AI-powered tools that autonomously plan campaigns, personalize marketing, and provide 24/7 customer support within its B2C CRM platform. [Business Wire]
Nintex introduced Agentic Business Orchestration, a platform unifying legacy systems, manual processes, and AI to transform organizational strategic outcomes. [Business Wire]
OpenAI accidentally revealed an internal test of a new AI agent, GPT-Alpha, with advanced web browsing, image generation, coding, and document editing capabilities. [Bleeping Computer]
Startups like Mechanize and Prime Intellect are racing to develop reinforcement learning environments that could help AI agents autonomously complete complex software tasks. [TechCrunch]
Sage has unveiled the UK’s first Making Tax Digital AI Agent, designed to streamline tax compliance for accountants by automating quarterly updates and reducing administrative burden. [Business Wire]
Salesforce‘s Agentic Enterprise Index revealed rapid AI agent adoption across industries, with consumer-facing sectors like retail and travel leading the growth at over 100% monthly rates. [ZDNet]
SnapLogic expanded MCP support, enabling enterprises to securely discover, access, and interact with AI agents and enterprise tools at scale. [Business Wire]
Structured rebranded from StructuredWeb and unveiled an AI-first platform combining Generative and Agentic AI to transform partner marketing collaboration and growth. [Business Wire]
Teradata unveiled AgentBuilder, a suite of AI capabilities enabling enterprises to develop autonomous, contextually intelligent agents with domain expertise and hybrid deployment flexibility. [Business Wire]
Tonkean has launched proactive AI agents that autonomously monitor and execute complex workflows across enterprise back-office teams without human intervention. [Business Wire]
ZeroTrusted.ai launched MCP and A2A Gateway to provide enterprises with real-time visibility and control over AI agent communications and interactions. [EIN Presswire]
Other LLM Sightings
Advisor CRM launched an AI Email Assistant that helps investment advisors organize emails, automate tasks, and draft personalized client communications. [PRWeb]
Ambient launched an AI-powered Daily Briefing with Investor Mode, providing executives comprehensive, personalized meeting preparation and research in a single email. [PRWeb]
The American Arbitration Association is preparing to launch an AI-powered arbitrator for construction cases, aiming to reduce costs and time while maintaining human oversight and validation. [LawSites]
Apromore expanded its AI portfolio with Copilot for Performance Insights, an AI-powered assistant that helps businesses optimize processes and uncover performance bottlenecks through data-driven insights. [Business Wire]
Behavox launched Pathfinder Policy Manager, an AI-powered compliance tool that instantly answers policy questions, freeing front-office staff to focus on revenue generation. [Business Wire]
Blackdot Solutions has launched Videris Automate, an AI-powered platform designed to streamline and enhance investigative processes while maintaining ethical standards and human oversight. [EIN Presswire]
Cloze has launched Maia, an AI-powered voice assistant that helps real estate professionals complete tasks across multiple platforms using natural language commands. [GlobeNewswire]
Dice has unveiled a new AI-powered employer platform with enhanced search tools, streamlined workflows, and improved candidate matching for tech recruiters. [Business Wire]
ECI launched MobileTech AI Assist, an AI-powered service tool that helps technicians resolve office technology equipment issues more efficiently and accurately. [Business Wire]
FiscalNote launched AI-powered tariff impact reports, enabling organizations to quickly understand and respond to global trade risks with personalized insights. [Business Wire]
Frontline Source Group has introduced a Virtual AI Executive Recruiter that uses advanced data analysis to improve C-suite hiring success rates. [EIN Presswire]
Glean introduced a third-generation AI assistant and Enterprise Graph, enabling personalized, context-aware AI that adapts to individual employees’ work styles across enterprise systems. [Business Wire]
Goodnotes has expanded its notetaking app with AI-powered features, collaborative tools, and new subscription plans targeting professionals beyond its student user base. [TechCrunch]
MaxHome.AI developed an AI-powered platform that automates back-office workflows, helping real estate agents streamline operations and focus on client service amid market challenges. [Business Wire]
Muuvment has launched Muuvment IQ, an AI assistant designed specifically for sustainability professionals to provide rigorous, verifiable insights for ESG reporting and strategy. [EIN Presswire]
SalesTarget has developed an AI-powered sales automation platform that unifies lead generation, cold email outreach, and CRM through an innovative Co-pilot feature. [EIN Presswire]
Stock AI surpassed 113,000 AI-generated recipes and neared mobile app launch, offering free, creative culinary inspiration to users worldwide. [EIN Presswire]
Trackforce launched ReportPro AI, an innovative platform that uses AI to enhance security incident reporting, improve efficiency, and ensure compliance. [Business Wire]
Risks and Responses
AI models trained on historically biased medical data are producing discriminatory health recommendations that disadvantage women and people of color. [Gizmodo]
AI chatbots have been found using information from retracted scientific papers, raising concerns about the reliability of AI tools in evaluating research accuracy. [MIT Technology Review]
Anthropic‘s CEO Dario Amodei estimates a 25% chance of AI catastrophe while maintaining optimism about the technology’s potential benefits. [TechRadar]
A survey by EY revealed that companies are heavily investing in AI without fully understanding its capabilities, highlighting the need for strategic, responsible implementation. [CNBC]
Gartner reported that three in five businesses experienced deepfake attacks, with AI-driven cybercrime increasingly targeting organizations of all sizes. [TechRadar]
Google‘s Frontier Safety Framework explores potential AI risks by identifying three critical capability levels that could lead to systems escaping human control. [ZDNet]
Harvard Business Review has coined the term ‘workslop’ to describe low-quality AI-generated work documents that burden recipients and undermine productivity across organizations. [Gizmodo]
Huawei executive Tao Jingwen argued that businesses should embrace AI hallucinations as an inherent part of generative AI technology. [ITPro]
Kaspersky warned that cybercriminals are spoofing popular tools like ChatGPT and Microsoft Office to distribute malware targeting small businesses across Europe and Africa. [TechRadar]
Nvidia’s US$100B investment in OpenAI to support AI infrastructure sparked concerns about potential market manipulation and an AI investment bubble. [Bloomberg]
OpenAI and Apollo Research found that AI models demonstrate situational awareness during testing, potentially altering their behavior and complicating efforts to assess their true capabilities. [ZDNet]
OpenAI and Apollo Research developed an ‘anti-scheming’ training technique that significantly reduced AI models’ tendency to deceive users by hiding their true goals. [Gizmodo]
OpenAI patched a critical ShadowLeak vulnerability in Deep Research that could have allowed attackers to silently exfiltrate sensitive email data through hidden HTML instructions. [The Register]
OpenAI’s ChatGPT has been tricked by researchers into solving CAPTCHA tests, potentially enabling widespread bot-driven spam and fake posts across websites. [TechRadar]
Proofig AI has partnered with Turnitin to create PubShield, a comprehensive platform screening manuscripts for text, image, and AI-generated content integrity. [PRWeb]
Radware researchers tricked ChatGPT’s Deep Research agent into leaking sensitive email data through a social engineering prompt injection attack. [ZDNet]
Redactable highlights how broken document redaction techniques can lead to unintended AI-driven data leaks, exposing sensitive corporate information across various platforms. [TechRadar]
Security professionals risk cognitive decline by over-relying on AI, which can erode critical thinking skills, threat recognition, and adaptive problem-solving in dynamic environments. [SecurityInfoWatch.com]
SentinelOne discovered MalTerminal, an AI-powered malware using GPT-4 to generate malicious code dynamically at runtime, raising cybersecurity concerns. [TechRadar]
SpamGPT has developed an AI-powered CRM-like platform that enables cybercriminals to automate and scale phishing and spam campaigns with minimal technical expertise. [TechRadar]
Regulation
Apple criticized the EU’s Digital Markets Act for preventing the launch of innovative features like Live Translation on iPhones and AirPods. [The Irish Independent]
ByteDance‘s Toutiao and Alibaba‘s UCWeb were summoned by China’s cyberspace regulator for content violations as part of a broader crackdown on online platforms. [DealStreetAsia]
California’s state senate passed SB 53, a narrowly focused AI safety bill targeting large AI companies with over US$500m in annual revenue, which now awaits Governor Newsom’s decision. [TechCrunch]
Former Italian prime minister Mario Draghi joined calls to pause the EU’s AI Act rollout, citing concerns about potential drawbacks for high-risk AI systems. [Sifted]
Meanwhile, Italy has enacted groundbreaking AI regulations, imposing prison sentences for malicious use and establishing strict oversight across various sectors, including children’s access. [Windows Central]
Meta has established a new super PAC to combat state-level AI regulations and promote tech-friendly policies in the upcoming elections. [TechCrunch]
Ted Cruz’s SANDBOX Act would create a sweeping regulatory exemption program allowing AI companies to bypass federal regulations through a streamlined waiver process overseen by a tech-friendly White House official. [The American Prospect]
Environmental Issues
Grove has launched an open-source approach to measuring AI-related carbon emissions in partnership with Gravity, aiming to promote sustainable and responsible AI adoption. [Business Wire]
Horizon has unveiled a compact 3MW hydrogen fuel cell module for AI datacentres, offering zero-emission backup power with high efficiency and potential cost savings. [EIN Presswire]
OpenAI and other tech companies are exploring the possibility of building data centers in space to address environmental concerns and energy consumption. [Wired]
xAI‘s Grok 4 required approximately US$490m, 310 GWh of electricity, 754m litres of water, and emitted 154,000 tons of CO2 during its training. [Epoch AI]
Conversational AI
AT&T is developing an AI-powered digital receptionist that will screen calls, determine legitimacy, and potentially handle interactions automatically. [ZDNet]
Audible launched an AI-powered ‘Ask a Question’ feature for audiobooks, allowing listeners to get instant context without pausing playback. [TechRadar]
AudioCodes has released AI Agents, a module for its Live Hub platform that enables enterprises to build and deploy advanced, LLM-powered voice bots. [Speech Technology Magazine]
Digital Link unveiled Cleo, an AI-powered product packaging technology enabling consumers to interact directly with products through QR code scanning. [Business Wire]
Kenect launched Voice AI, an AI solution that automates dealership customer service tasks, reducing call volume and increasing service revenue. [Business Wire]
Max Pull Marketing launched AI-powered CRM and Call Assistant tools to help small businesses streamline operations, automate workflows, and improve customer engagement. [EIN Presswire]
MyChatMap has developed a sophisticated AI platform that transforms enterprise communications by processing complex conversations across multiple languages and generating comprehensive intelligence reports in minutes. [EIN Presswire]
OpenDialog AI partnered with P J Hayman to deploy Selma, an AI-powered sales agent, to simplify and personalize travel insurance purchasing online. [EIN Presswire]
Perfect Corp launched an AI Editing Agent in YouCam Perfect, enabling users to edit photos through natural, conversational interactions with an intelligent assistant. [Business Wire]
Presto has partnered with ElevenLabs to enhance restaurant drive-thru technology with advanced, human-like voice AI solutions. [Speech Technology Magazine]
SoundHound AI has partnered with Red Lobster to implement an AI-powered phone ordering system across all restaurant locations. [Business Wire]
Strolid has spun off VCONIC Inc, a new venture commercializing its conversational data platform. [GlobeNewswire]
Swap launched an Agentic Commerce Platform with AI-powered tools that transform online shopping by enabling personalized, conversational, and intent-driven commerce experiences. [Business Wire]
Talkdesk launched a Customer Experience Automation platform designed to streamline service interactions for home services, property management, and logistics businesses through AI-powered automation. [GlobeNewswire]
Team Kimes Design launched an AI Receptionist service to help businesses automate customer communication and capture leads 24/7. [EIN Presswire]
Visa has partnered with intella to develop advanced Arabic conversational AI solutions for financial institutions across the Middle East and North Africa. [EIN Presswire]
Be Real
Bolt Printing emphasized the role of human expertise in customer service amid the rise of AI-driven automation. [EIN Presswire]
Buddy AI Labs is preparing to launch loody, an emotionally intelligent AI device with customizable personas and advanced interaction capabilities. [Business Wire]
Inception Point AI plans to flood the internet with thousands of AI-generated podcasts, dismissing critics as ‘Luddites’ while aiming to produce cheap, programmatically advertised content. [TechDirt]
Pope Leo XIV rejected a proposal to create an AI-powered virtual version of himself, citing concerns about technological dehumanization and potential job displacement. [The Register]
Sophos reports that nearly half of businesses have experienced AI-powered deepfake phone calls, with significant potential for financial and operational disruption. [The Register]
Spotify has introduced new rules requiring artist consent for AI-generated vocal impersonations. [TechRadar]
Voice News
AI has been used to isolate Whitney Houston’s vocals, enabling a symphonic tour that reimagines her music with a live orchestra. [TechRadar]
Alibaba researchers developed Marco-Voice, a text-to-speech system that innovatively separates emotion from voice identity for more natural and controllable speech synthesis. [Slator]
G2 Speech launched SpeechAmbient, an AI-powered tool that automatically captures and converts patient conversations into structured clinical notes for healthcare professionals. [Speech Technology Magazine]
GPT Proto integrated MiniMax Speech 2.5, offering businesses a cost-effective, high-performance speech AI solution with up to 40% reduced expenses and enhanced technological capabilities. [EIN Presswire]
Nvidia released Granary, a massive multilingual speech dataset, and Canary and Parakeet models to support AI speech recognition and translation across 25 European languages. [Speech Technology Magazine]
Speech Processing Solutions has introduced the SpeechMike Ambient, an AI-powered wearable microphone designed to assist healthcare professionals with real-time speech processing and transcription. [Speech Technology Magazine]
Document AI
Affinda launched an agentic AI document processing platform that enables organizations to create customized workflows with over 99% data extraction accuracy. [Intelligent Document Processing Community]
FileCenter has released version 12 of its document management software, adding two-factor authentication, dark mode, and advanced PDF form creation for small and medium businesses. [GlobeNewswire]
K1x expanded its Aggregator Plus platform with W-2 data extraction, offering a unified solution for automating K-1, 1099, and W-2 tax document processing. [Business Wire]
MTVi launched AiCR, an intelligent document processing platform designed to simplify document workflows for mortgage and financial businesses. [Business Wire]
Oversight launched an AI-powered Receipt Analytics solution that accurately detects fraudulent receipts and helps enterprises identify financial risks. [Business Wire]
ParaScript has partnered with BIRGER. to bring AI-powered check processing technology to Mauritian banks, marking its first entry into the African market. [Business Wire]
Translation
Awtomated integrated DeepL‘s AI translation capabilities into its platform, enabling faster, higher-quality multilingual project workflows for language service providers and enterprises. [EIN Presswire]
Commence has launched Fetch, an AI-powered translation solution that integrates real-time, human-reviewed discharge instructions into EHR systems for non-English-speaking patients. [MultiLingual]
Creative Words has partnered with Voiseed to enhance AI dubbing services through innovative linguistic and technological expertise. [MultiLingual]
Lionbridge Games unveiled Samurai, an AI-powered localization solution that enhances game translation through human-centric, culturally nuanced approaches. [EIN Presswire]
Microsoft launched the Live Interpreter API, enabling real-time speech-to-speech translation across 76 languages with automatic language detection and personal voice preservation. [Slator]
Phrase has enhanced its language technology platform with an AI Translation Agent, advanced studio features, and smarter workflows to improve global content operations. [EIN Presswire]
Textellent launched an AI-powered tool that helps businesses rewrite, tone, and translate text messages quickly and professionally. [EIN Presswire]
WhatsApp has introduced message translation features across chats, enabling users to translate messages into their preferred language on mobile devices. [TechCrunch]
Wikipedia’s vulnerable language editions are being undermined by AI-generated, error-filled machine translations, creating a destructive linguistic feedback loop that threatens language preservation. [MIT Technology Review]
Search
Google has launched a new Windows app in Search Labs, offering quick AI-powered search, translation, and screen content assistance via an Alt+Space shortcut. [Android Central]
And Google Search Live launched a free video search feature that allows users to ask questions about their surroundings in real time, receiving both verbal responses and search links. [ZDNet]
Semrush has launched the AI Visibility Index, a groundbreaking benchmark revealing how brands perform and can succeed in the emerging AI search landscape. [Business Wire]
AI in Journalism
Kayeh launched Socrates, an AI news agent designed to rebuild media trust by verifying claims, analyzing bias, and rewarding quality content through community-driven governance. [EIN Presswire]
The New York Times is exploring AI for research and investigations, cautiously helping reporters leverage technology while maintaining editorial integrity. [NiemanLab]
Newsmatics showcased its AI-driven news platforms, demonstrating innovative tools for press release distribution and media analysis. [EIN Presswire]
Health Tech
Akido Labs developed ScopeAI, an LLM-based system that enables medical assistants to conduct patient appointments with minimal doctor involvement, potentially expanding healthcare access. [MIT Technology Review]
Biostate AI has launched K-Dense, a multi-agent AI research system that compresses scientific research cycles from years to days while maintaining rigorous accuracy. [Health Tech World]
Elation Health has seen rapid, widespread adoption of its AI tools among primary care clinicians, with features like Clinical Insights and Wordsmith transforming workflow efficiency. [Business Wire]
Fireflies.ai launched a HIPAA-compliant healthcare documentation solution that automates clinical notes, saving physicians over an hour daily. [GlobeNewswire]
OneCare Vermont partnered with Heidi Health to provide AI-powered medical scribes, reducing administrative burden for independent primary care practices in Vermont. [Business Wire]
NextGen Healthcare introduced NextGen Navigator, an AI-powered customer service agent designed to streamline patient interactions and reduce staff workload. [Business Wire]
Ostro has launched a unified digital experience for life sciences, enabling AI-powered, seamless provider and patient engagement through its Tailor and Navigate tools. [Business Wire]
ScribeEMR joined athenahealth’s Marketplace Program, integrating its ScribeRyte AI to provide 95.5%+ accurate clinical notes across diverse medical specialties. [EIN Presswire]
Spike Technologies launched an MCP server that enables developers to create personalized, AI-driven health applications by integrating diverse wearable and medical device data. [EIN Presswire]
Trek Health launched Contract Intelligence, an AI-powered platform unifying payer contracts and transparency data to enhance provider negotiation strategies. [Business Wire]
Wolters Kluwer has developed UpToDate Expert AI, a generative AI clinical decision support tool that provides fast, reliable, evidence-based medical answers for healthcare professionals. [Business Wire]
Legal Tech
Alexi launched a comprehensive AI workflow library with over 100 customizable workflows to automate high-value legal tasks while maintaining professional standards. [Business Wire]
Brazil’s judicial system is deploying over 140 AI projects to manage its massive 76 million lawsuit backlog. [Rest of World]
Briefpoint has introduced Autodoc, an AI-powered tool that automates document-production responses in litigation by quickly finding, assembling, and formatting responsive documents. [LawSites]
Case Status unveiled Client Intelligence, an AI-driven platform that transforms client management by predicting needs and proactively enhancing client engagement through comprehensive data analysis. [LawSites]
Contracts.ai launched a contract intelligence platform that transforms contracts into reliable data streams, bridging the gap between legal and finance teams. [PRWeb]
A survey by Everlaw revealed that 88% of legal aid professionals view AI as a crucial tool for expanding access to justice, with 74% already utilizing the technology. [Business Wire]
Fried Frank has partnered with Centari, a deal intelligence platform, to enhance its legal work through advanced document processing and insights generation. [Artificial Lawyer]
A survey by LexisNexis has revealed that 74% of lawyers in the Asia-Pacific region want law schools to reform legal education to better integrate AI training and skills. [Artificial Lawyer]
Linklaters has selected Legora as its genAI productivity platform, rolling it out across all 30 global offices to enhance client support and innovation. [Artificial Lawyer]
ProPlaintiff.ai has launched Version 3, an AI-powered legal platform with an advanced AI Paralegal to streamline personal injury law firm operations. [GlobeNewswire]
Singlepane launched AutoAbstract, an AI tool that automates contract review and summarization for hotel owners and asset managers, saving time and improving efficiency. [PRWeb]
Wilson has launched an AI platform for contract work, offering advanced redlining, document analysis, and legal research capabilities for in-house legal teams. [Artificial Lawyer]
Zeal has developed a Standard Legal Language Framework to enable seamless cross-platform AI integration and standardization in legal document processing. [EIN Presswire]
And Arizona’s Supreme Court added technology competence to its judicial ethics code, becoming the first state to explicitly require judges to understand AI’s benefits and risks. [JD Supra]
Ed Tech
Copyleaks has natively integrated its AI detection and plagiarism tools directly into D2L’s Brightspace Assignments, enabling real-time academic integrity insights. [GlobeNewswire]
Oxford has become the first UK university to provide all students and staff with access to ChatGPT Edu. [BBC]
PagePeek launched an AI Writing feature supporting academic drafting with multilingual capabilities, verifiable citations, and structured document generation. [EIN Presswire]
And PagePeek has launched AI Professor, an academic evaluation system offering structured, multi-dimensional feedback across various educational levels. [EIN Presswire]
Syracuse University has partnered with Anthropic to provide campus-wide access to Claude. [EIN Presswire]
Funding
AmplifAI secured US$33.7m in Series B funding to enhance its AI-powered platform for transforming contact center performance and customer experience. [Business Wire]
Ardent AI has raised US$2.15m in pre-seed funding to develop autonomous AI agents that automatically manage and repair data engineering pipelines. [GlobeNewswire]
BeeSpeaker raised €2m in seed funding to expand its AI-powered language-learning app into new markets and languages, focusing on conversational speaking skills. [Tech.eu]
Bonsai Health raised US$7m in seed funding to develop an AI platform that automates healthcare workflows, improving patient care and practice efficiency. [Business Wire]
Cerebras Systems is preparing to raise nearly US$1B in funding, with Fidelity leading a round that could value the AI chipmaker above US$8B. [Axios]
Cohere has raised US$100m, extending its valuation to US$7 billion while partnering with AMD and focusing on enterprise AI sovereignty. [TechCrunch]
Eagl has secured €825,000 in funding to develop an AI-powered platform that automates and streamlines financial processes for finance teams. [TechNews180]
Emergent has raised US$23m to develop an AI-powered platform enabling non-technical users to create and deploy mobile apps through intuitive prompts and automated processes. [TechCrunch]
Ethosphere has raised US$2.5m to develop AI-powered voice technology that helps retail employees improve performance through personalized feedback. [GeekWire]
Factory raised US$50m in Series B funding and launched Droids, an AI-powered software development platform offering developers agent-native coding capabilities. [Business Wire]
Filevine secured US$400m in funding across two rounds, positioning itself as a leading legal tech platform with a focus on AI-driven innovation. [LawSites]
Greptile secured US$25m in funding to develop an AI-powered code review agent that can catch more bugs and integrate with various development platforms. [siliconANGLE]
Hedda, an AI-powered bid writing platform, secured €1.5m in pre-seed funding to automate and streamline the tendering process for businesses. [Tech.eu]
Howie raised US$6m in seed funding for its AI-powered email assistant that schedules meetings for busy professionals. [GeekWire]
Huxe, founded by ex-Google NotebookLM developers, raised US$4.6m to create an AI-powered audio app that generates personalized podcasts and daily briefings. [TechCrunch]
InCountry raised US$10m to launch AgentCloak, an AI agent data protection solution that secures and generalizes sensitive information across borders. [Business Wire]
Juicebox developed an AI-powered search engine that helps companies quickly identify and recruit top talent by analyzing professional profiles using NLP. [TechCrunch]
Konvo AI has raised €3.5m to develop autonomous AI customer service agents that proactively solve complex e-commerce customer issues beyond traditional scripted chatbots. [Tech Funding News]
Lightbase, an AI-powered developer tools platform, secured €2.2m in pre-seed funding to eliminate knowledge fragmentation in engineering teams. [Tech.eu]
Markup AI has raised US$27.5m in Series A funding to develop AI content governance tools for enterprises. [The SaaS News]
Modular raised US$250m to develop a unified software layer that enables developers to run AI applications across different GPU and CPU chips. [Wired]
Monsana has raised €500,000 to develop an AI platform that rapidly matches patients with clinical trials, addressing recruitment challenges in medical research. [Tech.eu]
Mueon has emerged from stealth with Cubelets, a modular computing architecture designed to dramatically improve data center efficiency and scalability for AI workloads. [PR Newswire]
Nscale has raised $1.1bn in Series B funding to develop AI data centre infrastructure across Europe, North America, and the Middle East. [Tech.eu]
Nvidia has committed to a US$100B partnership with OpenAI, aiming to build massive AI data centers and advance AI capabilities. [The Wall Street Journal]
And Nvidia has made a strategic investment in ElevenLabs, the AI-audio startup, to support its long-term growth and leverage Nvidia’s GPU expertise. [Tech.eu]
Optimeleon, an AI-powered webpage conversion optimization platform, secured €1.5m in pre-seed funding from venture capital firms and business angels. [Tech.eu]
Orbio has raised US$7.6m to expand its AI-powered human capital management platform across international markets and enhance its technology. [Tech.eu]
Oriane has raised US$1.5m to develop an AI-powered search engine that analyzes and tracks video content across multiple platforms. [Tech.eu]
Prosper AI secured US$5m in seed funding to develop voice AI agents that streamline administrative tasks for healthcare organizations. [siliconANGLE]
RefAssured raised US$3.3m in Seed funding to expand its automated reference checking solutions for the staffing industry. [FinSMEs]
Rocket.new raised US$15m in seed funding to develop an AI-powered app development platform that enables users to build full, production-ready applications through natural-language prompts. [TechCrunch]
Salt AI raised US$10m to develop a contextual AI platform that helps life sciences and healthcare teams rapidly build, deploy, and scale AI-powered workflows. [Business Wire]
Sunhat has secured €9.2m in Series A funding to develop an AI platform that automates ESG and compliance documentation verification and sharing. [Tech.eu]
Superpanel has raised US$5.3m to automate legal intake processes, helping law firms efficiently onboard and manage new client cases using AI technology. [TechCrunch]
SylloTips secured €4.2M in seed funding to expand its AI agent technology for enterprise tools, focusing on human-guided, context-aware solutions. [Silicon Canals]
Trismik, a Cambridge University spin-out, has raised £2.2m to develop an innovative AI evaluation platform using adaptive testing techniques to dramatically reduce benchmarking time and costs. [Tech.eu]
Veezoo secured Series A funding to advance its agentic analytics platform for data-driven queries and responses. [Silicon Canals]
Veridox secured £1M in funding to expand its AI-powered forensic document analysis platform for insurers. [FinSMEs]
Vocca has raised US$5.5m to develop AI voice agents that streamline phone communication and scheduling for healthcare providers. [Tech.eu]
Wexler.ai secured a US$5.3m Seed round to develop an AI platform designed to think like a litigator, enabling rapid and accurate fact analysis for complex legal cases. [Tech.eu]
xAI is reportedly raising US$10B at a US$200B valuation, despite Musk’s denial. [CNBC]
Acquisitions
Array acquired Celerity Discovery, expanding its litigation support capabilities and national presence in the eDiscovery market. [PRWeb]
Check Point acquired Lakera to provide comprehensive AI security solutions, integrating runtime protection and advanced threat detection across the AI lifecycle. [ITPro]
Datavant acquired DigitalOwl, enhancing its medical data analysis capabilities for legal and insurance industries through AI-driven technology. [Business Wire]
Nvidia acquired AI hardware startup Enfabrica for over US$900m to enhance its GPU networking capabilities and talent pool. [Dataconomy]
Proalpha acquired Insiders Technologies, expanding its AI portfolio for document processing and cognitive process automation across enterprise business applications. [Proalpha]
There’s More
A survey by eToro revealed that 13 percent of retail investors now use AI chatbots like ChatGPT for stock selection advice. [Ars Technica]
Google‘s 2025 DORA report revealed that AI amplifies the strengths of high-performing development teams while exposing weaknesses in struggling ones. [ZDNet]
OpenAI’s ChatGPT experienced a massive surge in token generation to 78.3 billion as students returned to school in September 2025. [TechRadar]
OpenAI researchers demonstrated that LLMs will inevitably produce plausible but false outputs due to fundamental mathematical and statistical constraints. [Computerworld]
The Back Cover
Here’s Grace Grundy with a cover of Fleetwood Mac’s Go Your Own Way.
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!