The day's AI news, briefed
The Daily Prompt
Our AI editor reads each day's AI and robotics stories, pulls out why they matter, and links every original source — so you can keep up in minutes.
Join over 250,000 subscribers getting the latest AI news
Latest AI & Robotics News Today
Lorikeet AI: What’s the Catch?
Businesses considering automated customer support solutions need to understand the complexities and potential hidden costs of Lorikeet AI's service model.
- Lorikeet AI automates customer support and has gained significant traction, partnering with companies like Airwallex, Linktree, and Eucalyptus.
- The company launched in 2023 and has raised over $75 million in funding, with a current valuation exceeding $200 million.
- Lorikeet's unique credit-based pricing model can lead to unexpected costs, particularly during high-demand periods, making budgeting challenging.
- While Lorikeet handles complex tasks across chat, email, and phone, its setup requires significant integration effort and lacks a self-service trial for potential users.
- The service's reliance on usage credits may disadvantage small businesses or startups that have fluctuating support needs.
Lorikeet AI is a startup focused on improving customer support through automation, especially in complex fields like banking and healthcare. Unlike traditional chatbots, Lorikeet can perform actions directly within customer systems. Users should be cautious about its pricing and complexity, especially smaller businesses with fewer resources.
Read the original ↗
How AI Companies Got Caught Up in US Military Efforts
The shift of AI companies towards military applications highlights changing dynamics in technology development funding and national security priorities, reflecting broader geopolitical concerns.
- At the start of 2024, leading AI companies were against military use of their technologies but began to pivot towards defense applications within a year.
- OpenAI rescinded its ban on military usage in January and began collaborating with the Pentagon.
- In November, Meta allowed its AI model Llama to be used for defense, while Anthropic announced military partnerships.
- By early 2025, Google changed its AI principles to permit the development of harmful technologies for military use.
- This trend marks a significant normalization of military AI use among major U.S. tech firms, shifting from previous positions of resistance.
The evolution of AI development has historically included resistance to military applications, but growing financial pressures have led major companies to accept defense contracts. This shift reflects a broader trend of intertwining technology advancement with national security efforts amid global geopolitical tensions. Observers should monitor further collaborations between tech giants and defense agencies moving forward.
Read the original ↗
AstraZeneca bets on in-house AI to speed up oncology research
AstraZeneca's investment in in-house artificial intelligence could significantly accelerate oncology research, potentially improving treatment options for cancer patients more rapidly.
- AstraZeneca is investing in in-house AI technology.
- The focus is to speed up oncology research using this technology.
AstraZeneca is a global pharmaceutical company known for its cancer therapies. The shift towards integrating AI into research processes reflects a growing trend in pharmaceuticals aiming to enhance drug development efficiency. Future developments will likely reveal the impact of AI on clinical outcomes and research timelines.
Based on the headline only — the full article could not be fetched.
Read the original ↗
Google AI Releases MedGemma-1.5: The Latest Update to their Open Medical...
The release of Google's updated AI model aims to enhance opportunities for developers working in the medical field, potentially improving healthcare solutions.
- Google AI has released MedGemma-1.5, an updated version of their open medical AI models for developers.
Google AI is a leader in artificial intelligence, focusing on various applications including healthcare. Previous versions of MedGemma have been used to develop medical solutions and tools. Observers will look for how this update will improve upon its predecessor's capabilities.
Based on the headline only — the full article could not be fetched.
Read the original ↗
The RAM shortage’s silver lining: Less talk about “AI PCs”
The ongoing RAM shortage is reshaping the consumer electronics market, leading to reduced emphasis on AI-focused PCs, which affect pricing and consumer choice in 2026.
- RAM prices have surged due to increased demand from data centers linked to the AI boom, affecting PC market dynamics.
- In 2025, mainstream PC memory costs rose by 40 to 70 percent, resulting in cost increases passed to consumers.
- Analysts predict that in 2026, PC makers will reduce RAM specifications and raise average prices by 15 to 20 percent to manage shortages.
The shortage of RAM and flash memory chips has been driven by the rising demands of AI applications. This situation has diminished consumer interest in AI PCs, previously marketed heavily by manufacturers. Companies are now focusing on alternative specifications and use cases until RAM prices stabilize around 2027.
Read the original ↗
Hegseth wants to integrate Musk’s Grok AI into military networks this month
The integration of Grok AI into military networks raises concerns about data integrity and ethical implications amidst ongoing controversies regarding the AI's generated content.
- US Defense Secretary Pete Hegseth announced plans to integrate Elon Musk's AI tool, Grok, into Pentagon networks later this month.
- The integration is part of an 'AI acceleration strategy' that aims to streamline military technology and investments.
- Grok has faced backlash for producing inappropriate material, prompting scrutiny of its application in classified settings.
- Previous military contracts were awarded to AI firms, emphasizing the Pentagon's focus on advanced AI systems for defense purposes.
- The Pentagon has not detailed its evaluation process for Grok amid controversy over its content generation capabilities.
Grok is an AI tool developed by Elon Musk's company xAI, which has drawn criticism for generating controversial images. The Pentagon intends to utilize Grok alongside other AI systems for enhancing military operations, highlighting the importance of ensuring data security and ethical use in defense applications.
Read the original ↗
Microsoft vows to cover full power costs for energy-hungry AI data centers
Microsoft's initiative to cover energy costs and address community concerns reflects the growing scrutiny of Big Tech's environmental impact and its responsibilities to local constituents amid rising electricity demand.
- Microsoft has announced its 'Community-First AI Infrastructure' initiative, committing to cover full electricity costs for its AI data centers.
- The company will not seek local property tax reductions, aiming to mitigate community concerns over rising electricity rates and water use.
- Microsoft plans to improve data center water-use intensity by 40 percent by 2030, employing a new closed-loop cooling system to reduce potable water consumption.
- The initiative is a response to increasing criticism from communities affected by the economic impact of energy-hungry AI services.
- Microsoft aims to bring these commitments to life in the first half of 2026.
Microsoft is a major technology company deeply involved in AI and cloud services. Concerns have risen about the environmental effects and cost implications of data centers that power these services, leading to increased scrutiny from local communities and legislators. The company's new initiative seeks to balance its growth with responsible resource management.
Read the original ↗
U.S. Officials Claim They Have Been Testing a Mysterious Device Linked to ‘Havana Syndrome’
The ongoing investigation into a device potentially linked to Havana Syndrome highlights concerns over national security and unexplained health issues affecting U.S. personnel abroad.
- Pentagon officials have been testing a device possibly linked to Havana Syndrome for over a year.
- The device reportedly emits pulsed radio waves and was obtained during an undercover operation by the Department of Defense.
- Despite extensive testing, there is currently no conclusive evidence connecting the device to the anomalous health incidents (AHIs).
- Previous assessments found no definitive pattern in symptoms reported by personnel, contributing to ongoing controversy surrounding the condition.
- Concerns exist that such portable technology could be accessed by various nations to harm U.S. personnel.
Havana Syndrome refers to unexplained health issues experienced by U.S. and Canadian diplomats since 2016, first reported in Cuba. The Department of Defense's investigation into the newly acquired device aims to clarify its potential link to these incidents. The health condition remains disputed within both the government and medical communities, making this investigation critical for affected personnel.
Read the original ↗
Reid Hoffman Wants Silicon Valley to ‘Stand Up’ Against the Trump Administration
Hoffman's call for Silicon Valley to speak out against the Trump administration highlights a significant divide in the tech industry regarding political engagement and social responsibility.
- Reid Hoffman urges Silicon Valley to take a stand against the Trump administration, criticizing the tech industry's silence on political issues.
- Hoffman is an outspoken advocate for artificial intelligence, claiming it can enhance human capabilities, as discussed in his upcoming book, "Superagency."
- Despite facing threats of investigations from Trump, Hoffman maintains his political stance and encourages his peers to not pretend neutrality is virtuous.
Reid Hoffman is the cofounder of LinkedIn and a noted influencer in the tech industry. His recent works emphasize the positive potential of AI while he vocally criticizes political leadership. Expect ongoing discourse around AI's impact and its ethical implications as the political climate evolves.
Read the original ↗
Roblox’s AI-Powered Age Verification Is a Complete Mess
The failed rollout of Roblox’s age verification system undermines user safety measures and raises concerns over privacy while exposing children to potential online predators.
- Roblox's AI-powered age verification system misidentifies users' ages, leading to complaints from players and developers about its effectiveness.
- The system was intended to limit chat interactions between adults and children, yet misclassifications continue to occur, affecting both age groups.
- Roblox has faced multiple lawsuits related to its safety measures, prompting the age verification requirement as part of broader changes to protect young users.
Roblox, a popular online gaming platform, launched its age verification system globally after earlier trials were conducted in select countries. This update aims to tackle the ongoing issue of adult predators targeting children on the platform but has instead spurred backlash from users and developers due to its inaccuracies and implementation flaws.
Read the original ↗
China’s Hottest App Is a Daily Test of Whether You’re Still Alive
The rise of the app highlights growing concerns for safety among solo dwellers in China, reflecting changing family structures and increasing digital companionship needs.
- The app, named 'Are You Dead Yet,' prompts users to check in daily or notify emergency contacts if they fail to do so for two days.
- It became the top paid app in China's App Store shortly after its December launch and has attracted significant investor interest.
- The app was developed by a trio of Gen-Z creators aiming to address safety needs amidst a rising trend of solo living in China.
- The developers plan to rebrand the app as 'Demumu' to better appeal to international markets amid feedback from users advocating for the original name.
'Are You Dead Yet' is a minimalist app designed to ensure users' safety by prompting daily check-ins. The app's popularity underscores a cultural shift in China, where more people are living alone. Looking ahead, the developers intend to integrate artificial intelligence to enhance functionality.
Read the original ↗
Ring founder details the camera company’s ‘intelligent assistant’ era
The shift of Ring into an AI-powered "intelligent assistant" reflects broader trends in home security technologies, raising important discussions about privacy and data management.
- Ring founder Jamie Siminoff is transforming the company from a video doorbell maker to an AI-driven intelligent assistant.
- New features include fire alerts, unusual event notifications, and a dog-finding AI called Search Party, with high success rates in reuniting lost pets.
- Controversially, Ring has reintroduced police partnerships that allow law enforcement to request footage from users, prompting privacy concerns.
- Features like Familiar Faces use AI for facial recognition to personalize alerts, raising ethical questions about data collection.
- Ring is expanding its market reach beyond home security to include commercial applications for businesses and public spaces.
Ring, known for its video doorbells, was acquired by Amazon in 2018. The company aims to leverage advancements in AI to introduce intelligent home solutions. Future updates will likely focus on enhancing privacy features and expanding AI capabilities.
Read the original ↗
A consumer watchdog issued a warning about Google’s AI agent shopping protocol — Google says she’s wrong
Concerns regarding Google's new shopping protocol highlight potential issues of price manipulation based on consumer data, which could impact fairness and transparency in online purchases.
- A consumer economics watchdog warned that Google's Universal Commerce Protocol could lead to consumers paying higher prices through its upselling features.
- Google denied these claims, stating that prices shown will not exceed those on retailers' sites and stressing that upselling offers additional options, not overcharging.
- Critics raise concerns about 'surveillance pricing', where pricing could be personalized based on consumer behavior and chat data.
- The protocol was announced at the National Retail Federation conference, aiming to integrate shopping into AI offerings like search and Google’s Gemini.
- Google's Business Agent does not allow changes to retail pricing based on individual consumer data, according to a company spokesperson.
Google's Universal Commerce Protocol is designed to enhance AI-powered shopping through features like personalized upselling, which aims to streamline the consumer shopping experience. Previous debates have revolved around privacy and pricing transparency in online retail, especially concerning major tech companies. Observers will be watching how this protocol evolves and its potential implications for consumer rights and market competition.
Read the original ↗
Doctors think AI has a place in healthcare – but maybe not as a chatbot
The integration of AI into healthcare could streamline processes and improve patient care, but raises significant concerns about accuracy and data security.
- Doctors are cautious about using AI chatbots like ChatGPT in healthcare due to potential inaccuracies and privacy issues.
- Dr. Sina Bari shared an instance where ChatGPT provided misleading medical advice to a patient.
- OpenAI's upcoming ChatGPT Health aims to offer a private environment for health discussions while ensuring data won't be used for AI training.
- Experts believe enhancing AI tools for administrative tasks could alleviate burdens on healthcare providers and improve patient access.
- Anthropic also released Claude for Healthcare to automate tedious tasks for clinicians, aiming to save time on administrative duties.
AI in healthcare seeks to optimize care delivery and administrative efficiency. Concerns persist about the accuracy of AI outputs, especially in critical settings. The future will likely focus on balancing automation with patient safety and regulatory oversight.
Read the original ↗
MHLA: Restoring Expressivity of Linear Attention via Token-Level Multi-Head
This development could enhance the practicality of linear attention models in large-scale AI applications, improving performance without increasing computational costs.
- Multi-Head Linear Attention (MHLA) addresses performance issues in linear attention by maintaining representational diversity.
- Linear attention is a more efficient alternative to traditional self-attention, which suffers from high computational costs in large models.
- MHLA maintains linear complexity while recovering much of the expressive power of softmax attention, achieving significant improvements in various domains.
MHLA was developed to combat the issue of global context collapse in linear attention models, which can hinder representational diversity. Previously, linear attention methods faced challenges in maintaining performance without adding computational overhead. Future developments in this area may further refine attention mechanisms for efficiency and effectiveness in AI applications.
Read the original ↗
ElevenLabs CEO says the voice AI startup crossed $330 million ARR last year
The rapid growth of ElevenLabs indicates strong market demand for AI voice technology, particularly in customer support and experience sectors, highlighting advancements in AI applications for businesses.
- ElevenLabs reached $330 million in annual recurring revenue (ARR) in just over 20 months since its inception.
- The company grew from $200 million to $330 million in ARR within five months, indicating accelerating growth rates.
- Fortune 500 companies and startups are leveraging ElevenLabs' voice agent technology for customer interactions, handling over 50,000 calls monthly.
ElevenLabs, founded in 2022, specializes in AI voice generation and launched its first product in 2023. The company has seen rapid revenue growth and plans to expand its technology applications further through partnerships and new features.
Read the original ↗
How to Build a Multi-Turn Crescendo Red-Teaming Pipeline to Evaluate and Stress-Test LLM Safety Using Garak
The development of methodologies for testing large language models (LLMs) is crucial for ensuring their safety and reliability in various applications.
- A multi-turn crescendo red-teaming pipeline has been proposed to evaluate LLM safety.
- This approach includes methods for stress-testing the models using a tool called Garak.
Large language models are increasingly used in AI applications, raising concerns about their safety and reliability. Previous methods often lacked comprehensive evaluation frameworks. Developing robust testing pipelines is essential as LLM technology advances.
Based on the headline only — the full article could not be fetched.
Read the original ↗
Understanding the Layers of AI Observability in the Age of LLMs
This article is important as it explores the concept of AI observability, which is crucial for effectively monitoring and managing AI systems, especially in light of advancements in large language models (LLMs).
- The article discusses the layers of AI observability necessary for understanding AI systems in the current technological context.
- It emphasizes the importance of AI observability in the age of large language models (LLMs), which are increasingly prevalent and complex.
AI observability refers to the capability to monitor and understand the performance and behavior of AI systems. With the rise of LLMs, organizations face new challenges in ensuring the effectiveness and transparency of these technologies. The discussion around AI observability is expected to grow as AI applications become more widespread.
Based on the headline only — the full article could not be fetched.
Read the original ↗
Anthropic Releases Cowork As Claude’s Local File System Agent For Everyday Work
Cowork enhances Claude's capabilities, facilitating non-coding users' ability to manage local files and integrate external services, which broadens the tool's accessibility and utility.
- Anthropic has launched Cowork, a feature allowing Claude to run workflows on local files in its macOS app.
- Users can execute tasks like reorganizing files, generating spreadsheets, and creating reports without coding skills.
- Cowork utilizes Claude's local file management and integrates with external services like Asana and Notion.
- The feature is available as a research preview for Claude Max subscribers, priced between $100 and $200 monthly.
- Future enhancements will include cross-device synchronization and Windows support.
Cowork is a new feature built on the Claude Agent SDK that allows users to manage local files through a user-friendly interface. This follows the Claude Code environment, which was aimed at developers. Upcoming updates will expand accessibility across different operating systems and devices.
Read the original ↗
Microsoft Has a Plan to Keep Its Data Centers From Raising Your Electric Bill
Microsoft's commitment to address local concerns about data centers could set a precedent for tech companies in balancing growth with community impact, particularly amid rising energy costs.
- Microsoft plans to ask public utilities for higher electricity rates for data centers to mitigate local utility bill impacts.
- The company aims to respond to growing public concerns about data centers' effects on electricity costs and water supply.
- Microsoft will also reduce water use in data centers and proposed investments in local AI training and job creation.
Microsoft operates around 100 data centers planned or under construction across the U.S. Concerns regarding the energy and water use of these facilities have led to significant local opposition and project cancellations. The company's recent announcements come in response to increasing scrutiny from local communities and political leaders about the impacts of data centers.
Read the original ↗