← Back to home

Daily briefing · Auto-updated

AI News

Top stories from across the AI world, curated daily.

Last updated: 23 August 2026, 07:00 AM IST
6 stories today
Bloomberg August 22, 2026
Nvidia has notified its biggest customers that AI server systems built around the upcoming Vera Rubin and Grace Blackwell chips will cost 15% or more above current prices when they ship early next year, driven by soaring memory costs

The hikes vary by chip generation and memory configuration, with HBM supply tightening at the leading edge of production. For hyperscalers already spending tens of billions on GPU clusters each quarter, another price floor reset compounds the cost of building out AI infrastructure.

Bloomberg August 22, 2026
Apollo Global’s chief economist finds AI is cutting wages, not jobs: workers in high-AI-exposure roles earned 6.7% less than their peers post-2023, costing 5.8 million U.S. workers $28 billion in annual pay while employment headcounts stayed flat

Service workers took the sharpest relative hit at 24% below comparable low-exposure roles. The study tracked 321 occupations from 2015–2025 and found the effect shows up as wage compression rather than the layoffs that most AI-impact forecasts predicted. Apollo notes AI is simultaneously fueling record business formation, which could eventually offset the wage drag.

OpenAI August 20, 2026
OpenAI is rolling out advertising in ChatGPT across 31 European markets on August 24 — six months after the US launch — with targeting limited to conversation context and location, and explicit GDPR consent required for any personalization

Free and Go tier users see ads; Plus, Pro, Enterprise, Business, and Education accounts remain ad-free. Opting out changes which ads appear but does not eliminate them. OpenAI framed the consent architecture as GDPR-compliant by design, with ad VP Dave Dugan confirming no behavioral profiling beyond the current session.

CNBC August 15, 2026
Anthropic posted over $11.5 billion in Q2 2026 revenue — up 14-fold year over year and more than double Q1 — while recording its first adjusted operating profit, two years ahead of its own internal schedule

The company attributed the surge to rapid enterprise adoption of Claude and Claude Code. At this run rate Anthropic approaches $46 billion in annualized revenue, materially reshaping the economics of the frontier AI industry and validating the thesis that AI infrastructure spending eventually converts to software margin.

Bloomberg August 12, 2026
Lovable, the Swedish vibe-coding startup that lets users build apps in plain English, raised $400 million at a $13.3 billion valuation as its annualized revenue approaches $600 million — with a headcount of just 146 people

The Series C was led by Menlo Ventures and the EU’s Scaleup Europe Fund, more than doubling Lovable’s December valuation of $6.6 billion. Launched in November 2024, the company plans to scale to 450 employees and deepen agentic workflow and multi-agent orchestration capabilities with the new capital.

TechCrunch August 11, 2026
Anthropic began embedding invisible, machine-readable watermarks in all text generated by new Claude models worldwide — not just in Europe — using Google DeepMind’s SynthID technique, triggered by the EU AI Act’s August 2 enforcement deadline

The watermark persists through copy-paste and survives light editing; Anthropic plans to release a public detection API. EU AI Act Article 50 requires generative AI systems launched after August 2 to include such marks, but Anthropic opted for global rollout rather than a geofenced implementation, effectively exporting the EU’s transparency standard everywhere Claude operates.