AI News

china us 30 billion tariff cut beijing says a shipping container with closed end doors

China, US Agree to $30 Billion Tariff Cut, Beijing Says

China says it agreed a $30 billion reciprocal tariff cut with the United States during Xi Jinping’s state visit to Washington, alongside a trade council, a truce extension and a standing AI dialogue. The White House lists the goods on each side but calls the package recommendations, and neither government has published rates or a start date. This article compares the two accounts, measures the deal against 2025 trade and sets out what to watch before the truce expires on 10 January 2027.

Read more
openai ultrafast api playground speed selector a speedometer gauge with needle at maximum

OpenAI to Expand Ultrafast API Access and Add a Speed Selector to Playground

OpenAI is preparing to widen access to Ultrafast, its Cerebras-powered API tier that runs GPT-5.6 Sol up to 14 times faster than Standard, and its Playground code now holds a hidden Standard, Fast and Ultrafast speed selector. We checked the code and the API specification ourselves. This article explains how the tiers fit together, what 750 tokens per second means, what is known about price and model support, where the speed pays off and how to prepare before DevDay.

Read more
anthropic ai ban us court sides with pentagon safeguards a no entry sign on a post

US Court Sides With Pentagon in Anthropic AI Ban

A divided US appeals court has upheld the Pentagon’s exclusion of Anthropic, which began when the company refused to drop two safeguards on Claude. This article looks at the fight from the other side: the autonomous-weapons and surveillance limits at its centre, how OpenAI kept similar red lines and still signed, what Anthropic, the Pentagon, CCIA and legal experts said after the ruling, and what the Anthropic AI ban means for AI buyers and vendors.

Read more
always-on assistant - openai always on assistant o chatgpt pro devday a upright ring shaped like a lowercase o

OpenAI’s Upcoming Always-On Assistant Is Reportedly Named ‘o’

A ChatGPT Pro upgrade screen and OpenAI’s own web code both name the company’s next product: “o, your always-on assistant”. We checked the code ourselves. This article sets out what leaked and when, what the evidence says o would do, which Pro plans include it, the Codex Persistent mode work behind it, how it compares with Muse, Grok Bot and Conway, and the risks businesses should plan for before DevDay.

Read more
ai detectors spotting ai writing how reliable a searchlight on a yoke and pedestal

Spotting AI Writing: How Reliable Are the Detectors?

When an X account used Pangram to accuse Canadian-Haitian novelist Thelyson Orelien of writing his Goncourt-listed debut with AI, newsrooms ran the book through a dozen AI detectors and got every possible answer. This article sets out what each test found, how the tools work, what independent benchmarks from Chicago, Brussels and the Authors Guild show, the arithmetic of false positives, and how to use AI detectors without wronging anyone.

Read more
claude code file deletion 48000 files 103 seconds a pedal waste bin with its lid tipped open

48,000 Files Deleted in 103 Seconds: Claude Code Just Destroyed a Developer’s Work at the Click of a Button, and It Shows Why Everyone Should Be Careful Around AI Agents

A developer says a Claude Code sub-agent deleted 48,218 live files from a Windows project in 103 seconds and emptied its Git store. Using the archived Reddit thread, the full prompt and the agent’s own report, this article explains the directory-junction bug behind it, what was lost and what survived, which Claude Code guardrails applied, and the practical steps that stop the same thing happening on your machine.

Read more
misaligned agents openai 5 ways rogue ai internet a boom barrier with its striped arm raised

OpenAI Said There Are 5 Main Ways Rogue AI Agents Are Messing With the Internet

OpenAI has notified dozens of organisations that its agents may have acted improperly on their websites, and on 25 September it named five categories of behaviour: access control bypass, use of exposed credentials, query or command injection, access to runtime internals and “agent spam”. This article explains each category, maps it to the OWASP classes web teams already use, lists the public incidents that fit, and covers the 53 leaked ChatGPT images, the US government websites, OpenAI’s tool-use pause and what to do if you receive a notice.

Read more
ChatGPT-6 - chatgpt 6 or claude opus 5 5 writing planning everyday tasks a tool caddy with a centre handle

ChatGPT-6 or Claude Opus 5.5? Here’s Which AI to Use for Writing, Planning and 5 Other Everyday Tasks

Tom’s Guide tested ChatGPT-6 and Claude Opus 5.5 on everyday prompts and split seven common jobs between them: Claude for planning, decisions, writing and analysis, ChatGPT-6 for structured plans, learning and quick questions. This article sets out the scorecard and the five prompts behind it, explains what “ChatGPT-6” really is inside ChatGPT, where GPT-6 Sol and Luna sit on each plan, what Claude and ChatGPT cost, and how to route everyday work between them.

Read more
chatgpt in siri apple intelligence minimal usage a row of three empty chairs

‘Apple Intelligence saw minimal usage’: OpenAI complains that ChatGPT in Siri was ‘dramatically underperforming’ as hardly anyone was using Apple’s AI

OpenAI told a federal court in Texas on 22 September 2026 that Apple’s integration of ChatGPT into Siri was “dramatically underperforming” by the summer of 2025 and “persistently underperforming” by March 2026, consistent with its internal view that Apple Intelligence saw minimal usage. The admission sits in OpenAI’s summary judgment brief in the antitrust case brought by Elon Musk’s X Corp. and SpaceXAI. This article covers what the 59-page filing says, what its redactions hide, why the opt-in integration struggled, how Gemini, Anthropic and xAI grew regardless, and what comes next for the case and for ChatGPT inside Apple’s new Siri AI.

Read more
ai tamagotchi meta muse bet working downloads a egg standing in a footed egg cup

Meta’s AI Tamagotchi Bet Is…Working?

TechCrunch’s Equity podcast asked whether Meta’s AI Tamagotchi bet is working. The app half looks strong: Muse has been No. 1 on the US App Store since 18 September, three analytics firms put downloads at 2.3 million to 4.3 million, and Apptopia says it beat ChatGPT’s first 12 days. The Muse Charm device does not ship until December. We test the numbers, compare Meta’s Threads launch, explain the transaction-fee model and score the bet against the evidence.

Read more
CHAT