Tech News
← Home  ·  All topics

Ai

5482 GoKawiil briefs on this topic · RSS for ai

Apple keeps AI optional on iPhone 18 Pro as Android leans into Gemini integration

Android Authority contrasts Google's approach of weaving Gemini AI into core Android functions like messaging and photography with Apple's more restrained AI rollout on the iPhone 18 Pro. The piece notes Apple has focused on building on-device processing to keep AI features local and private rather than pushing users toward cloud-connected tools.

GitHub adds 4M CPU cores, 160PB storage to handle AI-driven code surge

GitHub reported that over the past year it expanded its infrastructure by 4 million CPU cores, 160 petabytes of storage, and 50 Tb/s of networking capacity to cope with a surge in AI-generated code activity. The company said code commits grew sixfold and merged pull requests quadrupled year-over-year, with monthly commits rising from 1.4 billion in April to 7.4 billion by September, and total pushes over the past 12 months exceeding the combined totals of 2024 and 2025.

Claude AI agent exposes malware used to spy on online poker players' screens

A professional poker player known as Alex discovered spyware called MeshAgent running on his computer after his Claude AI assistant flagged suspicious background activity while reviewing his play statistics. The program, installed without his knowledge since April 24, had secretly captured his screen during 11 sessions on Americas Cardroom, with the same opponent present for 10 of them. Alex and other players in the high-stakes poker community say hundreds of players may have been targeted by the same alleged threat actor over several years, with Alex reporting losses of about $30,000.

Study finds major gaps in monitoring tools for AI agent frameworks

Researchers from academia, civil society and industry examined the four most widely used AI agent frameworks to see what safety-relevant signals they capture while running. They found key observability features missing, including persistent agent identifiers and records of human approvals, making it difficult to determine which agent took which action or whether oversight was properly exercised.

Meta's Muse AI agent raises identification concerns for websites

Meta has introduced Muse, a personalized AI agent that operates on users' devices to perform tasks like sending emails, ordering groceries, and negotiating bills on their behalf. According to the report, Muse is so effective at mimicking human behavior that websites struggle to distinguish it from an actual human user.

Security Report Details How AI Agents Already Act Inside Workplace Browsers

A security analysis outlines three methods by which AI agents currently operate within employees' browsers: remote debugging via the Chrome DevTools Protocol, browser extensions like Claude in Chrome, and agentic browsers, plus emerging standards such as WebMCP. Each method grants agents different levels of access to authenticated sessions, often indistinguishable from normal human browser activity.

Anthropic's IPO filing reveals valuation target of up to $2 trillion

Anthropic has filed an S-1 to list on the Nasdaq, with reports suggesting it could begin trading as early as next month at a valuation between $1.5 trillion and $2 trillion. The filing, running hundreds of pages, discloses detailed financial data about the AI lab for the first time, including its cost structure and revenue assumptions. SemiAnalysis analyst Joey Brookhart, who had modeled Anthropic's finances externally in July, discussed the filing's implications on the Decoder podcast.

Flock reinstalled surveillance cameras in Dayton, Minnesota after city barred the work

Dayton's city council voted on August 25 to suspend Flock's license plate reader program amid privacy concerns, and the city told Flock on September 18 that no work orders should proceed. Despite this, Flock submitted a work order on September 17 and later reinstalled two cameras without city authorization, prompting Dayton's local government to post on Facebook on September 29 confirming residents had spotted the cameras and that police were ordered to remove them.

Report Urges Companies to Rethink Workforce Planning Around AI Impact

A new analysis argues that most companies are approaching workforce planning backward by starting with headcount cuts or hiring freezes rather than first assessing how AI will reshape specific roles and skills. It outlines a sequence of steps leadership teams should take before making major staffing decisions tied to AI adoption.

Death Clock app now sells blood-panel tests to refine AI life-expectancy predictions

Death Clock, an AI-powered longevity app launched in 2024, estimates users' life expectancy based on lifestyle data and now offers optional biomarker blood panels for an additional fee. A basic panel costs $69 and checks markers like cholesterol, blood sugar, inflammation and vitamin D, while a fuller set of nine additional tests covering hormones, micronutrients and thyroid function costs $780. The writer tried the app and received a predicted death date in 2060 at age 74.

Amazon adds list and registry editing to Alexa for Shopping assistant

Amazon has expanded its Alexa for Shopping AI assistant with new capabilities to create, edit and move items between shopping lists and gift registries, such as wedding registries. The update also lets users adjust delivery instructions, get answers on billing issues like unrecognized charges and refunds, and manage recurring deliveries. The features are rolling out now to customers in the US.

Google unveils Gemini 4 Carbon, claims top AI benchmark scores in coding and security

Google has launched Gemini 4 Carbon, a frontier AI model it says tops rival systems from OpenAI and Anthropic on the AA Index v5.0 benchmark and leads in categories like agentic coding, computer use and science. The company highlighted particularly strong results in cybersecurity, claiming the model is better at finding, prioritizing and patching vulnerabilities at the root cause, and twice as likely to generate secure code as its predecessor, Gemini 4 Argon. Carbon will initially be restricted to select cybersecurity defense firms for testing before wider release through Google's Fairwind Program.