All articles tagged with #ai agents
OpenAI is sunsetting Atlas—the AI-enabled browser launched nine months ago—after learning from user behavior; the move accompanies updates like ChatGPT 5.6 and a new desktop app, ChatGPT Work, which extends Atlas’s capabilities with AI agents to handle tasks locally and on the web, with Atlas slated to shut down around August 9.
OpenAI has secured a broad rollout of GPT-5.6 after regulatory clearance and introduced ChatGPT Work, a ChatGPT–Codex hybrid that lets non-technical users generate documents, spreadsheets, presentations, and web apps by pulling context from apps, files, and workflows and connecting to tools via a unified plugins directory. Access starts on desktop for most users with mobile/web rollout in stages; the move aims to outpace rivals in the AI agent space and offers a lower-cost option within the GPT-5.6 lineup (Sol, Terra, Luna).
Anthropic expands Claude Cowork to smartphones and the web, letting the task‑automation agent keep running without an active desktop. The rollout starts in beta for Max plan subscribers (from $100/month), with potential later access for Pro and possibly free tiers, enabling cross‑app data work (email, Slack, transcripts) to draft documents and emails after hours. This mirrors a broader shift toward always‑on mobile AI agents and pits Anthropic against OpenAI and Google, though risks like prompt injections remain a concern.
Apple’s Doug Brooks says the Mac mini (and Mac Studio) are in “incredible demand” for running AI agents locally, as developers favor systems that are always on and under user control. He frames agentic AI as a whole‑chip challenge rather than just a GPU one, powered by Neural Engine and other neural accelerators to enable on‑device AI. Brooks envisions a hybrid future where some workloads run locally while others are offloaded to the cloud for privacy and cost reasons, and cites “transparent AI” features on iPhone/iPad and Mac (like Draw Things and SwingVision) as examples of AI quietly working in the background. The interview underscores Apple’s chip‑and‑software approach to push on‑device AI forward, a trend likely to accelerate after WWDC 2026.
Meta CEO Mark Zuckerberg said at an internal town hall that AI agent tech hasn’t progressed as quickly as hoped, and the sweeping restructuring that cut about 10% of the workforce and reassigned roughly 7,000 employees to AI-focused roles hasn’t yet borne the expected fruit; he cited optimism about tools like Anthropic’s Claude Code and said Meta could spend up to $145 billion on AI infrastructure this year, with meaningful benefits expected in 3–6 months.
OpenClaw has released standalone AI agent apps for iOS and Android, enabling users to chat with its assistant and grant broad device permissions (camera, screen, location, photos, contacts, calendar, reminders). The project is open-source and run by a foundation after founder Peter Steinberger joined OpenAI, with OpenAI reportedly providing some support; the move follows ongoing tension with Apple over agentic AI tools.
Edison Scientific, spun off from FutureHouse with roughly $70 million in funding, is being tapped by the Metsera-backed team and Population Health Partners to scale the creation of new biotech companies using AI agents capable of autonomous reasoning, reflecting pharma’s growing interest in AI-driven drug discovery.
AI startup Taktile closed a $110 million Series C led by Goldman Sachs with Tiger Global, Index Ventures, and Y Combinator to turn AI models into dedicated agents for sensitive financial tasks, such as processing insurance claims and expediting payouts. The funds will further build out the platform and support a new São Paulo office.
Apple is reportedly exploring its own OpenClaw-like AI agents that could run across iPhone, iPad, and Mac, potentially bundled with Apple One or offered via a Services-focused model; while this could advance Siri and automation, Apple would face security and privacy challenges that could limit agent access and delay rollout, per Bloomberg's Mark Gurman.
Ben Thompson argues that while Microsoft pitches Solara as an enterprise cloud-based agent ecosystem, Apple doubles down on the iPhone as the central AI hub, with Siri AI featuring personal-context understanding and cross-app actions via App Intents and Spotlight. Apple’s approach relies on on-device intelligence complemented by private cloud compute, aiming for useful, privacy-conscious assistance for consumers who mainly seek entertainment rather than productivity; in this view, Apple’s iPhone-centric strategy may outpace vaporware-focused enterprise visions.
Microsoft CEO Satya Nadella says AI agents should be treated like human employees, with assigned identities, sandboxed access, and auditable policies. He highlighted the governance challenges of managing roughly 100 AI coding agents at once and noted Microsoft’s Agent 365 suite (including Entra and Purview) is built to improve security, containment, and observability as the company scales AI use.
Microsoft CEO Satya Nadella publicly rejected an internal memo that proposed making users addicted to the Scout AI agent, saying the company’s AI strategy should empower people and contribute real value and broad economic growth, not encourage addiction; the memo’s leak has spotlighted AI engagement tactics.
Cloudflare data show AI agents and bots now generate 57.4% of web requests, surpassing human traffic at 42.6%, signaling a rapid surge in autonomous browsing. The rise—driven by AI agents using tools with little human input—could reshape the internet’s business model since bots don’t click ads, prompting new monetization ideas and raising questions about how content is discovered and consumed in the web’s evolving landscape.
Meta internal documents show the company considering pricing Hatch, its planned consumer AI agent, at up to $199.99 per month with a premium tier for higher usage; pricing isn’t final, but the move signals paid AI assistants that could compete with OpenAI and Anthropic offerings.
Morgan Stanley will enable corporate clients’ autonomous AI agents to connect directly to its stock-plan platforms ShareWorks and Equity Edge, allowing data access and insights without human logins, aiming to scale plan administration and the wealth-management funnel; early access has been granted to some clients, with broader rollout planned for next year, using the Model Context Protocol to plug AI models into data sources.