AI News

space 2 pro anker ai chip soundcore headphones a retro radio blank dial handle v2

Anker Brings Its AI Chip to the Soundcore Space 2 Pro Headphones

Anker used IFA 2026 to spread its Thus compute-in-memory AI chip across five audio models, led by the $200 Soundcore Space 2 Pro over-ear headphones shipping September 22. The chip claims 150 times more AI compute than Anker’s previous flagship silicon while eliminating the energy cost of moving data, powering an 8-sensor adaptive ANC array, real-time voice separation on calls, and HearID 5.0 personalised sound over a Hi-Res/LDAC path. This guide covers the full spec sheet, the Liberty Buds 2, AeroClip 2 and 2 Pro, sleep earbuds and OTC hearing aids around it, and what on-device AI audio means for buyers.

Read more
catch ai admin assistant calendar email phone calls a desk telephone handset cradle

Catch Launches AI Admin Assistant for Calendar, Email, and Phone Calls

Catch AI opened to the public on 2 September 2026 as an AI admin assistant that executes rather than suggests — working the inbox, running the calendar, and placing real phone calls in an executive’s name at $99 a month. The launch lands alongside a $5 million seed co-led by Entrée Capital and Pitango. This article covers the phone-calling capability that separates it from the field, the channels and integrations it runs on, the free Uta scheduling concierge, the travel beta, and what a working AI admin assistant means for executives and lean teams.

Read more
gemini 3 8 flash agent studio google cloud a gemini 3.8 flash console slider tracks

Gemini 3.8 Flash Available in Agent Studio on Google Cloud Platform

Gemini 3.8 Flash landed in Agent Studio on Google Cloud Platform on 2 September 2026 — Google’s third Flash release in six weeks. It scores 73.7% on DeepSWE v1.1, within 0.3 points of Claude Opus 5, at $0.75 per million input tokens until January. This article covers what the model changes inside Agent Studio, the benchmarks and pricing worth trusting, the restricted Flash Cyber security variant, and what a three-weekly release cadence means for teams building production agents.

Read more
forward-deployed engineer - forward deployed engineering how enterprise ai learns a solid ridge tent field camp

Forward-Deployed Engineering Is How Enterprise AI Learns

Forward-deployed engineering has become the centre of the enterprise AI land grab. AWS is spending $1 billion to embed thousands of engineers with customers, Palantir’s decade-old playbook is now the industry default, and OpenAI and Anthropic hire for the role by name. This article explains what forward-deployed engineers actually do, why 95% of AI pilots failed without them, how the embedded learning loop works, what the talent shortage means, and how mid-sized businesses can buy the same loop at their own scale.

Read more
wonderful agent builder autonomous agents claude a potters wheel with pot v2

Wonderful Agent Builder Creates and Refines Autonomous Agents Using Claude

Wonderful, the enterprise agent platform, launched Agent Builder — an autonomous AI agent powered by Anthropic’s Claude that creates, tests and refines other AI agents end to end. With build times down up to 50%, early production issues down 20%, more than 60 enterprise deployments and a $2 billion valuation reached within a year, this article unpacks how the Wonderful Agent Builder works, why Claude powers it, and what agents that build agents mean for enterprise AI teams.

Read more
alexa for shopping scam message verification a large magnifying glass round rim

PSA: Amazon’s Shopping AI Can Now Tell You if That Message Is a Scam

Amazon’s AI shopping assistant just gained a scam checker. Ask Alexa for Shopping whether an email, text or phone call really came from Amazon and it checks the message against billions of records, returning a verdict in seconds. With 360,000 customers a year asking Amazon “is this real?” and the FTC reporting a record $3.5 billion lost to imposter scams in 2025, this article unpacks how the verification feature works, the scams it targets, how to use it, and what it means for shoppers and businesses.

Read more
hiddenlayer 100m ai security funding a large upright shield

HiddenLayer Nabs $100M as Enterprises Rush to Secure Their AI Deployments

AI security startup HiddenLayer has closed a $100 million Series B led by Delta-v Capital, with Ten Eleven Ventures, Morgan Stanley, Microsoft’s M12 and Booz Allen Ventures participating. The round follows a year of 10x revenue growth, 50+ new enterprise customers and deepening US defence work, and lands as Gartner says AI security spending will near-triple to $4.78 billion by 2027. This article unpacks the deal, the platform, the adversarial threats behind the demand, and what the raise means for any business deploying AI.

Read more
agentic 7g roadmap ai powered wireless networks a hot air balloon solid basket

A Roadmap for Agentic 7G AI-Powered Wireless Networks

HKUST researchers led by Khaled B. Letaief have published the first engineering roadmap for agentic 7G wireless networks in npj Wireless Technology. Their reasoning-empowered task-oriented communication framework lets AI agents decide what to share, when to transmit and why — turning networks from bit pipes into engines of collective intelligence. This article unpacks the three core capabilities, the 3GPP Release 21 timeline it sits against, the five open research gaps, and what agent-first networking means for businesses today.

Read more
acer rtx spark ai desktop future a mini desktop tower two front slots v3

Acer Shows Off Its AI Desktop Future With the RTX Spark Design

Acer used IFA 2026 to show off the SFF RTX Spark, a small-form-factor desktop concept built on NVIDIA’s RTX Spark superchip: up to a 20-core Grace CPU, a 6,144-core Blackwell GPU, 128 GB of unified memory and one petaflop of AI compute in a vertical shell with a fold-out stand. Alongside it, the palm-sized Veriton RI110 mini workstation ships first, from Q4 2026. This article breaks down the specs, the local agentic AI pitch, the vendor ecosystem shipping from fall 2026, and what a petaflop on the desk means for creators, developers, gamers and business buyers.

Read more
pangram ai detection gold standard trust a solid rubber stamp thick handle

Pangram Has Emerged as the Gold Standard of AI Detection. Should You Trust It?

Pangram, a 24-person Brooklyn startup, has become publishing’s de facto AI referee: its scores cancelled a Hachette novel, embarrassed The New York Times and cast doubt on a major literary prize, and Substack now bakes the detector into its platform. Independent NBER benchmarking really does put it far ahead of GPTZero, Originality.ai and Turnitin on false positives — yet humanizer tools evade it over 96 percent of the time, verdicts wobble on short excerpts, and bias questions hang over who gets accused. This article weighs the evidence from WIRED’s investigation, the Chicago Booth benchmark and the company’s own disclosures to answer the practical question: when should you trust a Pangram score, and when should you not?

Read more
CHAT