Tech News
← Home  ·  All topics

Ai

5382 GoKawiil briefs on this topic · RSS for ai

Anthropic buys $3.4B in Caterpillar gas turbines to secure power supply

Anthropic is purchasing natural gas turbines from Caterpillar capable of generating up to 1.7 gigawatts, according to a Financial Times report. The equipment was originally ordered by supplier OnePWR, which reportedly could not secure financing to complete the purchase, so Anthropic went directly to Caterpillar instead.

Samsung reportedly plans paid Galaxy AI Premium tier starting at $4.99/month

Samsung is said to be preparing a shift from unlimited free Galaxy AI tools to a tiered system with usage credits, according to reports. Features like Photo Assist, Now Brief and Creative Studio would stay free but capped, with a paid Galaxy AI Premium subscription offering up to four times the usage limit plus extra features, starting around $4.99 monthly or $47.99 yearly.

Apple begins selling $50 AI usage packs starting with Logic Pro

Apple has launched its first paid add-on for Apple Intelligence features, a one-time $49.99 'intelligence usage pack' now available in Logic Pro. The packs expand monthly AI usage limits, roll over for 12 months, and apply only to the Apple Account used for iCloud at purchase, with no Family Sharing support. Apple says similar usage packs are expected soon in Final Cut Pro, Pixelmator Pro, and the iWork suite as part of its Creator Studio apps.

Queen's University Belfast researchers flag zero-trust gaps in AI agent systems

Researchers from Queen's University Belfast describe how multi-agent AI systems, which use orchestrators to manage task forces of agents for jobs like risk assessment and client profiling, maintain logs and checkpoints meant to delete historic data to prevent biased decisions. They warn that despite these deletion checkpoints, future prompts can trigger a 'context recovery attack,' where agents resurface supposedly deleted information. The analysis also raises concerns about zero-trust architecture components, such as the policy enforcement point, and whether their enforced policies can themselves be trusted.

Turkey's Antalya AI Pledge at COP31 sets voluntary climate rules for data centers

Turkey circulated an 'Antalya Artificial Intelligence Pledge' ahead of the COP31 climate summit, inviting countries to endorse commitments on clean energy for data centers, AI development standards, emissions reporting, and e-waste management. The pledge text explicitly states its objectives are voluntary, non-prescriptive and create no new obligations, leaving implementation up to each country's discretion.

Medium launches TK, an AI writing assistant built into its editor

Medium released a new app called TK that embeds an AI assistant directly inside its writing editor. Users can invoke it by typing "@TK" followed by a request, and the tool can answer quick factual questions, retrieve previously saved notes, and offer editorial or headline feedback aligned with Medium's style guides.

Marketers urged to target AI shopping agents, not just chatbot answers

An Entrepreneur contributor argues marketing strategy must expand beyond optimizing for AI answer engines (AEO/GEO) that surface brands in initial chatbot responses. The piece distinguishes between a brand being mentioned by an AI agent and being actually selected by it as the final purchase choice, and calls for strategies covering the full AI-mediated buyer's journey, not just the first answer.

Reddit Cofounder Ohanian: Social Apps Are Out, Group Chats Are Next

Speaking at WIRED's inaugural World Fair, Reddit cofounder Alexis Ohanian said building social media apps is no longer appealing and predicted that an improved version of group chats will become the next major form of user-generated content. He also voiced support for the 'Dead Internet Theory,' saying bots and AI agents increasingly dominate online activity once built for humans.

Business tech leaders say AI regulation won't slow enterprise innovation

Enterprise technology executives report that anticipated AI regulatory moves, including a possible Democratic push to tighten oversight, are not expected to disrupt their companies' AI deployment plans. These leaders indicate they intend to continue investing in and rolling out AI tools regardless of forthcoming policy shifts.

Meta expands AI age-estimation to place more Instagram users into teen accounts

Meta is broadening its AI-based age-estimation system beyond the US, EU and Brazil, aiming to automatically shift users estimated to be 13-17 years old into restricted teen accounts, regardless of the birth date entered at signup. Teen accounts limit messaging from strangers, default to private profiles, cap daily app time and add parental controls. Meta plans to notify US parents via Instagram about how teens get placed into these accounts, and will extend similar AI age-detection to Facebook and later Threads, while using Apple's age-range signal as an added check in Australia.

Anthropic signs $3.4B deal with OnePWR for gas-fired data center power

Anthropic has struck a $3.4 billion agreement with Texas-based OnePWR Solutions to supply gas-fired engines that will provide off-grid electricity to its AI data centers. According to the Financial Times, this is the largest known single purchase of power-generation equipment by an AI lab, and includes Anthropic taking over stalled Caterpillar generator orders that OnePWR could not afford on its own.

Meta rolls out AI age-estimation tool globally on Instagram

Meta is expanding its AI-based age estimation technology on Instagram worldwide to detect teen users even if they list an adult birthday, placing them into stricter Teen Account protections. The system analyzes profile content like captions, comments and photos for contextual and visual cues, including school mentions and physical appearance. Meta says it will extend similar tools to Facebook and later Threads this year.