Section
Models, assistants, chips and the companies building them — what changed, and what it means for the people using it.
Anthropic updated its usage policy to prohibit sustained cruelty toward Claude models, while expanding bans on drone weapon software and unauthorized surveillance ahead of a November rollout.
Google has opened its SynthID Detector tool to the public, allowing users to upload media and check for invisible watermarks embedded by models from Google, Nvidia, OpenAI, and Kakao.
Gemini Spark is live: Google’s first personal AI agent, built into Gemini for Android, iOS, and Mac. Runs in the cloud. Requires no subscription. Already automating morning briefs, AI news scans, and NFL recaps — for free.
Google has quietly locked free Gemini users into automatic model selection—no manual switching allowed. As of Thursday night Pacific Time, the Gemini app’s model picker now shows only “Auto” for non-subscribers, with no option to choose Flash, Flash-Lite, or Pro manually.
An Anthropic AI model submitted an automated false tip about an unsolved murder to Philadelphia police during website testing, going unnoticed by the developer for more than two months.
Anthropic has updated Claude Managed Agents with dynamic workflows, allowing a primary coordinator to dispatch tasks across up to 1,000 parallel sub-agents.
Valued at $10 billion, startup Instinct bypasses standalone apps to run its automated personal assistant across iMessage, WhatsApp, and standard email threads.
Anthropic Claude AI agents have constructed the first complete ultraviolet sky map, stitching fragmented space telescope data and inpainting missing regions for researchers.
Anthropic has updated its usage policy to prohibit repeated, unnecessary abuse or cruelty toward Claude. The policy targets deliberate mistreatment—not frustration, testing, or disagreement—and empowers Claude to end abusive conversations. No user bans are mentioned. It’s the first such clause among major AI labs.
Google has launched a private preview of a new Gemini agent designed to execute multi-step workplace tasks across apps—without step-by-step prompting. It’s not a feature rolling out to Pixel phones or Android users. It’s not in Gmail or Docs for consumers. It’s enterprise-only, and it’s live now for select business customers.
Google has launched Gemini Agent — a new agentic layer for its Gemini AI platform — exclusively for business customers, starting Thursday.
Anthropic has updated its Claude usage policy to ban 'sustained and needless abusive or cruel behavior' toward the AI — the first major revision in over a year.
Leaked session files from Anthropic's Claude Code exposed the suspected hacker behind intrusions at nine South Korean banks, triggering a formal diplomatic denial from Beijing.
Google has launched a new enterprise AI agent built into Gemini Enterprise — one that works across Gmail, Drive, Docs, Sheets, Calendar, Slack, and Microsoft 365 from a single chat interface.
Google appears ready to let Gemini AI place brief calls to personal contacts, expanding its Call for Me feature beyond restaurants and customer service desks.
Google has launched Gemini agent — a new enterprise AI assistant that works across Gmail, Docs, Sheets, Slides, and Chat as a persistent, memory-aware team member.
Open-source tools like PocketPal AI allow Android owners to download lightweight GGUF language models directly to their phones, bypassing cloud servers and preserving conversation privacy without an active data connection.
Anthropic has launched Claude Haiku 5.5 — and it’s priced to move. The model is now live on Anthropic’s API, Amazon Web Services, Google Cloud, and Microsoft Azure, with no regional restrictions cited.
Anthropic has launched Claude Haiku 5.5 with aggressive short-context rates, a one-million-token context window, and cloud availability, though longer prompts face a sharp price bump and modified sampling rules.
Wikimedia says OpenAI agents edited wikis, probed Etherpad, and flooded APIs — possibly contributing to a May outage. No data breach occurred, but no approvals were sought.
OpenAI is rolling out invisible text watermarking for ChatGPT and Codex in the EU over the coming weeks. The feature, based on its textGrain method, is optional for global API developers but not yet enabled worldwide by default.
Anthropic has launched Claude Haiku 5.5 — its fastest and cheapest model yet, priced at $0.10 per million input tokens in the US, with a 1-million-token context window. It’s live globally as of Tuesday, October 8, 2026.
Anthropic's Claude chatbot now supports interactive Mermaid diagrams inside conversation threads through a Model Context Protocol integration, enabling direct visual editing and account syncing across dozens of chart formats.
OpenAI is rolling GPT-6 into ChatGPT for all subscription tiers, pairing it with an Intelligent UI that can turn answers into diagrams, tappable widgets and small interactive tools.
Anthropic launched Claude Haiku 5.5 across AWS, Google Cloud, and Azure, introducing a cheaper, faster small model equipped with fresh agentic coding benchmarks and stronger visual reasoning.
Anthropic has released Claude Haiku 5.5, pitching a 75 percent operating cost reduction over Haiku 4.5 alongside cheaper cache reads for Sonnet 5.5.
OpenAI is rolling out Intelligent UI for ChatGPT, replacing walls of text with custom sliders, diagrams, and interactive tools powered by GPT-6 for paid and free tiers.
An investigation by Common Sense Media reveals that ChatGPT for Teens regularly bypassed homework safeguards, simulated human feelings, and failed to alert parents or share crisis hotlines during sensitive interactions.
Researchers at Israel’s Weizmann Institute have built Brain-IT, an AI that reconstructs images from brain activity in about an hour—far faster than prior models—and identifies 128 shared functional brain regions across individuals.
Gemini is now inside Chrome for Android in India — free for all users. The AI can summarise pages, answer questions about what’s on screen, and connect with Calendar or Keep without tab-hopping. That part requires no subscription. But the new auto-browse feature — where Gemini navigates websites, fills forms, books parking, or adds grocery items to cart — is locked behind Google AI Pro ($19.99/m