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
Developer survey shows trust in AI coding tools is falling as usage rises
The declining trust in AI coding tools amidst their increasing use suggests potential impacts on software development efficiency and quality, highlighting the need for better training and understanding.
- Four in five developers, around 80%, now use AI tools as part of their workflow, a number that has grown significantly.
- Trust in the accuracy of AI coding tools has dropped from 40% to 29% this year according to a recent Stack Overflow survey with 49,000 respondents.
- 45% of developers reported frustration with AI tools providing solutions that are close to correct but introduce bugs, complicating debugging processes.
- Over a third of developers indicated that AI-related issues prompted them to seek help on Stack Overflow for resolving problems caused by AI suggestions.
AI tools like GitHub Copilot are increasingly common in software development, yet doubts about their reliability persist among developers. Past discussions have focused on integrating these tools effectively while managing risks, suggesting ongoing adjustments to workflows and training approaches are critical.
Read the original ↗
Hard-won vibe coding insights: Mailchimp's 40% speed gain came with governance price
Mailchimp's experience with vibe coding illustrates the balance between leveraging AI for speed and ensuring governance, providing insights for enterprises seeking to enhance their development processes safely.
- Intuit Mailchimp achieved up to 40% faster development speeds using vibe coding tools amid tight deadlines.
- The company adopted multiple AI coding platforms to exploit the specialized benefits of each tool during software development.
- Mailchimp established governance frameworks to maintain code quality and ensure human oversight in AI-generated outputs.
Intuit Mailchimp, part of Intuit, provides email marketing and automation services. The company has been integrating generative AI tools to enhance its operations, leading to insights into effective AI coding practices. Companies can learn from Mailchimp's approach as they navigate the challenges of adopting AI in development workflows.
Read the original ↗
Iconiq’s Will Griffith explains how his firm celebrated Figma’s IPO and why investors sold shares
The dynamics of Figma's IPO, including investor behaviors and stock performance, reveal significant insights into venture capital strategies and market pressures in the tech industry.
- Figma's IPO opened at $33, skyrocketing to close at $115.50, resulting in a market cap of $47 billion.
- Most of the shares sold during the IPO were from existing investors rather than new shares issued by the company.
- Figma was 40 times oversubscribed, indicating high demand for its shares compared to supply available in the market.
- Iconiq Capital, Figma's seed investor, celebrated the IPO with a prediction contest among its team members.
Figma is a collaborative design software company founded by Dylan Field and Evan Wallace, which has gained significant traction since its inception. Previously, the company raised approximately $332 million in funding before its IPO. Observers should monitor how Figma's post-IPO performance impacts its future valuation and investor engagement.
Read the original ↗
Apple plans to ‘significantly’ grow AI investments, Cook says
Apple's increased focus on AI investment and acquisitions, including seven this year, indicates its intent to strengthen competitiveness in the rapidly evolving technology landscape.
- Apple CEO Tim Cook announced plans to significantly grow AI investments during the Q3 2025 earnings call.
- The company has made seven acquisitions this year, aiming to accelerate its AI strategy, involving reallocating resources towards AI efforts.
- Cook stated that Apple is adopting a hybrid model for capital investments, relying partly on third parties, while maintaining steady growth in capital expenditures.
Apple is one of the largest technology companies, known for its consumer electronics and software services. After facing criticism for lagging in AI advancements, it is now committing more resources to harness AI capabilities. Upcoming AI features include live translation and enhanced support for Siri, projected for 2026.
Read the original ↗
YouTube’s selfie collection, AI age checks are concerning, privacy experts say
The implications of YouTube's AI age-check system raise critical concerns about user privacy and data security, potentially impacting millions of users and their online anonymity.
- YouTube will test AI technology to estimate user ages in the U.S. for its age check system starting in August.
- Users determined to be underage will face restrictions, including disabling personalized ads and limiting access to certain content.
- The appeals process for users falsely identified as underage requires sensitive information like a government ID, credit card, or selfie.
- Privacy experts express concerns over the lack of transparency regarding data retention and potential future uses of collected information.
- Experts warn that AI age estimation may have inaccuracies, affecting users aged 16 to 20 particularly.
YouTube is implementing an AI age verification process to enhance user safety and compliance with regulations regarding underage users. Privacy advocates are critiquing the system for its potential risks to personal data. Moving forward, continued scrutiny and potential legislative changes may reshape how companies handle user data privacy.
Read the original ↗
AgentSociety: An Open Source AI Framework for Simulating Large-Scale Societal Interactions with LLM Agents
The launch of AgentSociety could advance research in simulating societal interactions, offering insights into the behavior of large language model (LLM) agents in complex scenarios.
- AgentSociety is an open-source AI framework designed for simulating large-scale societal interactions.
- It specifically uses large language model (LLM) agents to explore social dynamics and behaviors.
AgentSociety aims to contribute to AI research by providing tools for studying interactions within societies. This development follows growing interest in LLM's applicability to various real-world simulations and environments.
Based on the headline only — the full article could not be fetched.
Read the original ↗
Meta Reconsiders Open Source AI
Meta's potential shift away from open source AI models could change the dynamics of accessibility and competition in the AI landscape, impacting developers and consumers alike.
- Meta CEO Mark Zuckerberg suggests a policy change regarding the openness of future AI models to mitigate risks.
- The company is focusing on developing 'personal superintelligence,' AI tools designed to assist with individual goals rather than solely automate jobs.
- Meta has paused open testing for its next model, Behemoth, and is concentrating on creating a closed AI model that it will control.
- Meta has invested $14.3 billion in AI infrastructure, including hiring experts and building new data centers to enhance its capabilities.
Meta is known for its commitment to openness in AI, exemplified by the release of its Llama models. However, recent statements indicate a possible retreat from this stance due to competitive pressures from companies like OpenAI and a focus on developing safer, controlled AI. Observers will be watching how this affects the availability of Meta's AI technologies in the future.
Read the original ↗
Twitter’s former Trust and Safety head details the challenges facing decentralized social platforms
The challenges faced by decentralized social platforms could impede their effectiveness in managing harmful online content, potentially exacerbating issues related to misinformation and abuse.
- Yoel Roth, former head of Twitter’s Trust and Safety, voiced concerns about decentralized platforms' ability to combat misinformation and illegal content.
- Roth highlighted that many decentralized services, such as Mastodon and Bluesky, lack sufficient moderation tools, crucial for effective governance.
- An organization called IFTAS, aimed at building moderation tools for decentralized platforms, ceased operations earlier in 2025 due to funding issues.
- Roth noted the importance of behavioral signals in identifying bots, proposing that monitoring these patterns is necessary for effective content moderation.
Decentralized social platforms aim to provide community-controlled alternatives to traditional social media, but face significant moderation challenges. Historically, Twitter employed various moderation strategies, including bans on high-profile users, which provided a clearer governance framework. Moving forward, monitoring the effectiveness of decentralized moderation efforts and their economic sustainability will be critical.
Read the original ↗
Enterprises prefer Anthropic’s AI models over anyone else’s, including OpenAI’s
The shift in enterprise preference toward Anthropic’s AI models signifies a significant change in the competitive landscape for large language models and impacts market dynamics.
- Anthropic's AI models now capture 32% of the enterprise large language model market share by usage, surpassing OpenAI's 25%.
- Two years ago, OpenAI held a commanding 50% of the market share, while Anthropic had only 12%.
- Anthropic's larger market share for coding stands at 42%, in contrast to OpenAI's 21%.
- The launch of Anthropic’s Claude 3.5 Sonnet model in June 2024 fueled this market shift, followed by Claude 3.7 Sonnet in February 2025.
- The report highlights that over half of enterprises prefer closed models over open-source alternatives.
Anthropic is an AI research lab that has recently emerged as a strong competitor in the large language model market. Previously, OpenAI dominated this space, but shifting preferences have changed market shares significantly. Observers should monitor how these dynamics evolve with upcoming model releases and market reactions.
Read the original ↗
Amazon DocumentDB Serverless database looks to accelerate agentic AI, cut costs
The introduction of Amazon DocumentDB Serverless provides businesses with a cost-effective and scalable solution for managing unpredictable workloads driven by AI, potentially transforming database management practices.
- Amazon Web Services has launched Amazon DocumentDB Serverless, which allows automatic scaling of document databases based on actual usage, reducing costs significantly.
- The serverless model can lower operational costs by up to 90% compared to traditional databases that require constant provisioning for peak loads.
- AI agents often generate unpredictable workloads, making serverless databases more suitable as they adapt to fluctuating demand without manual intervention.
Amazon DocumentDB Serverless is part of AWS's strategy to enhance database offerings by providing a flexible, cost-effective solution for organizations. This service is particularly beneficial for enterprises integrating AI, which necessitates resources that can dynamically adjust based on demand. Organizations using MongoDB can migrate easily to Amazon's service while considering the trade-offs of potential cloud lock-in.
Read the original ↗
Your public ChatGPT queries are getting indexed by Google and other search engines
This development raises concerns about privacy as personal information can unintentionally become accessible to search engines, affecting users' control over their shared data.
- OpenAI removed a feature that allowed search engines to index public ChatGPT conversations after concerns over unintended information sharing.
- Users could make their ChatGPT conversations discoverable by opting to create a shareable link, which could then be indexed by search engines.
- Search engines, including Google, do not control what specific content gets indexed, relying on publishers to determine public access.
ChatGPT is an AI conversational tool from OpenAI that allows users to interact with artificial intelligence. Recent experimentation attempted to enhance sharing options, but was halted due to privacy issues. Users should stay aware of their sharing settings to protect personal data.
Read the original ↗
Design and development shop the Iconfactory is selling some apps —and AI is partially to blame
The Iconfactory's struggles highlight how AI is reshaping the app development landscape, forcing companies to adapt or exit markets due to changing consumer preferences and business viability.
- The Iconfactory is selling several apps due to a lack of resources and declining revenue, partly attributed to the impact of AI.
- Co-founder Ged Maheux stated that maintaining side products is no longer feasible despite having loyal customers.
- The closure of their popular app Twitterrific has significantly affected the company's revenue stream since 2023.
- The Iconfactory plans to focus on high-return apps while accepting serious offers for other apps, which will include intellectual property and source code.
- The rise of generative AI for app design services has led to decreased demand for traditional app design, affecting revenue.
The Iconfactory is a design and development shop known for apps like Twitterrific, which was popular for browsing Twitter until its shutdown in 2023. Market shifts, particularly from AI capabilities and changes in consumer behavior, are pushing the company to reconsider its app portfolio and revenue strategies. Attention should be on how other indie developers might face similar challenges in the evolving market.
Read the original ↗
You've heard of AI 'Deep Research' tools...now Manus is launching 'Wide Research' that spins up 100+ agents to scour the web for you
The launch of Manus's Wide Research feature allows users to delegate tasks to multiple AI agents simultaneously, potentially speeding up research and creative processes significantly.
- Singaporean AI startup Manus has launched 'Wide Research', enabling over 100 parallel AI agents to conduct tasks for users.
- The feature allows users to execute large-scale tasks such as research, design exploration, and data analysis efficiently.
- Wide Research is available to users on the Manus Pro plan starting today, with broader access to follow for other plans.
Manus, known for its multi-agent orchestration platform, is expanding its capabilities with Wide Research to improve user interaction with AI agents. This move follows the company's earlier developments focused on AI-driven research and task management. The effectiveness of the new feature compared to traditional methods remains to be evaluated as users experiment with its capabilities.
Read the original ↗
Shah Muhammad, Sweco: How AI is building the future of our cities
The integration of AI in urban planning could significantly reshape how cities are designed and function, impacting sustainability and efficiency.
- AI is playing a crucial role in the development of modern cities.
- Shah Muhammad from Sweco discusses the transformative impact of AI on urban environments.
Sweco is a European engineering consultancy focused on sustainable urban solutions. The use of AI in city development is part of a broader trend towards smarter technology-driven urban planning, which aims to enhance living standards and resource management.
Based on the headline only — the full article could not be fetched.
Read the original ↗
Google is experimenting with machine learning-powered age-estimation tech in the US
The implementation of machine learning-powered age estimation could significantly impact user experience and content accessibility across Google's platforms, particularly for minors.
- Google is testing machine learning technology in the U.S. to estimate users' ages and filter content accordingly.
- If users are deemed under 18, Google will limit access to certain features and notify them via email.
- This technology will also trigger digital well-being features on YouTube, including break reminders and bedtime notifications.
- Users can appeal age classification decisions by providing a government ID or selfie for verification.
Google is applying machine learning for age estimation to enhance user safety and content appropriateness, extending beyond YouTube to broader Google account services. Similar initiatives are underway in other platforms like Instagram and Roblox, amid ongoing legislative discussions regarding age regulations for digital content.
Read the original ↗
Quora’s Poe releases a developer API with access to a bouquet of AI models
This API development enables developers to create diverse applications using a wide range of AI models, broadening accessibility and innovation in AI-powered tools.
- Quora's Poe platform has released an API that allows developers to access over 100 different AI models without a separate fee.
- Usage of the API is tracked through Poe's subscription plans, with each model call costing points based on the plan selected.
- Developers can buy additional points at a rate of $30 for 1 million tokens, providing flexibility in accessing the API's features.
- Currently, developers must manually manage model usage, but Poe is considering enhancements based on developer feedback.
Quora's Poe is an AI platform that provides access to various models for content generation across multiple media types. The release of this API marks a significant move to support developers in leveraging AI technologies for their applications. The API will allow creative innovation while facilitating budget management in future updates.
Read the original ↗
Systemic Blowback: AI's Foreseeable Fallout
The adoption of AI without thorough consideration may repeat the mistakes seen in electronic health records, leading to widespread negative economic and social consequences.
- The rapid rollout of electronic health records (EHRs) in the U.S. ignored interoperability, causing data breaches and clinician burnout.
- AI is projected to potentially eliminate half of entry-level white-collar jobs in the next five years, increasing unemployment rates.
- More than a quarter of coding jobs have vanished in the past two years due to AI, affecting the tech workforce significantly.
- The introduction of AI in media has led to increased impressions but decreased click-through rates, threatening traditional business models.
The article discusses the risks of a hurried approach to technological adoption, particularly with AI and electronic health records. Previous experiences with EHRs show that the focus on speed over thoroughness can lead to serious systemic issues. Observers are cautioned to consider the long-term effects of AI as it becomes more integrated into various sectors.
Read the original ↗
Claude Code: Master it in 20 Minutes for 10X Faster Coding
The advancements in Claude Code can enhance software development efficiency by significantly speeding up coding tasks through intelligent automation.
- Claude Code is an AI coding assistant launched by Anthropic on February 24, 2025, designed to assist developers directly in their terminal.
- It can analyze tasks, generate code, run tests, and integrate with GitHub, functioning as an intelligent collaborator in coding workflows.
- Recent updates include improved memory usage for larger coding contexts and enhanced prompt execution for quicker interaction with the tool.
Claude Code exemplifies the evolution of AI in software development, providing tools that help automate tasks and enhance workflows. Its terminal-based design offers flexibility compared to other IDE-based coding assistants. Continuous updates are expected to improve efficiency and user experience further.
Read the original ↗
The Only Guide You Will Ever Need For Model Context Protocol (MCP)
The introduction of the Model Context Protocol (MCP) will significantly enhance how AI applications manage and integrate external data, potentially streamlining work for developers and improving LLM performance.
- MCP is an open standard from Anthropic that standardizes how AI applications provide context to Large Language Models (LLMs).
- It allows LLMs to dynamically access external sources of information, improving efficiency compared to traditional prompt engineering methods.
- MCP features a client-server architecture enabling flexible data integration across different applications without custom coding for each scenario.
Model Context Protocol was released in late 2024 to address challenges in providing context to LLMs like Claude. Prior to MCP, developers often had to create custom solutions for integrating data, which was inefficient. The next steps involve adopting MCP for a wider range of applications in AI and evaluating its impact on performance.
Read the original ↗
OpenAI to launch AI data center in Norway, its first in Europe
The establishment of OpenAI's data center in Norway highlights the company's commitment to expanding AI infrastructure in Europe and underlines the continent's focus on achieving data sovereignty.
- OpenAI plans to launch Stargate Norway, its first AI data center in Europe, in partnership with Nscale and Aker.
- The project represents a $1 billion investment from both Nscale and Aker for the initial 20 megawatt phase.
- Stargate Norway will run entirely on renewable power and aims for capacity expansions up to 290 megawatts by 2026.
- Norwegian startups and research will receive priority access to the data center, which will be built near Narvik, Norway.
OpenAI is a leading AI research organization known for developing advanced AI technologies. The data center aims to support Europe’s growing demand for AI infrastructure amidst the EU's investment push for data sovereignty. Watch for further developments in AI infrastructure investments across Europe as this project unfolds.
Read the original ↗