Industry News — Page 10

Cloudflare's Workers AI platform now efficiently serves large Chinese frontier models like Kimi and GLM by quantizing the KV cache to FP8, doubling context window capacity and increasing throughput by 41%.

Mistral's Shieldstral, a 3B open-weights multimodal safety classifier released August 4, 2026, matches larger guardrail models on text safety benchmarks and sets a new state of the art for multimodal safety classification, covering 12 languages and running on a single 16GB GPU.

AMD's Q2 2026 earnings reveal record $11.5 billion revenue, with data center revenue more than doubling to $6.7 billion due to AI demand, now making up 58% of total revenue. This marks a significant shift for the company.

Palantir CEO Alex Karp criticized AI frontier labs as "Marxist" for allegedly using enterprise data to build competing businesses, following Palantir's record Q2 2026 revenue of $1.9 billion and profit of $1.1 billion.

Palantir CEO Alex Karp slammed AI frontier labs as "Marxist" after record Q2 2026 earnings, warning that using their models could fund competitors and urging enterprises to control their own AI "exhaust" with model-agnostic platforms.

Cloudflare's Workers AI platform efficiently serves large Chinese frontier models like Kimi and GLM using FP8 quantization, halving memory usage and doubling context capacity for 41% higher throughput and 30% lower cost per token.

Anthropic has signed a six-year, $10 billion computing deal with new cloud startup Volta Infra, the largest compute contract ever awarded to a company of its age. This agreement will provide Anthropic with access to advanced computing resources for its AI models.

The Trump administration's FTC banned foreign robot imports last week, citing national security and supply chain concerns. This aims to protect domestic robotics from Chinese competition and potential data risks from foreign-made humanoids.

Intelligence, creator of Design Arena, secured $7.9 million in seed funding led by Index Ventures. The platform uses human evaluation data from 5.3 million users to refine AI-generated designs and track regional taste differences.

Pippa AI is testing artist royalties to address the trust gap in ethical generative video, offering per-generation payouts and a share of subscription revenue to participating artists.

Fenix Flexin's "Rubberz" hit number 58 on the Billboard Hot 100, raising questions about AI involvement and the music industry's lack of policy for AI-generated tracks.

Windows 11 installation files now exceed 8 GB, more than doubling Windows 10's original size. This growth is primarily due to AI components for Copilot+ PCs, affecting storage for users and IT deployments.

Fenix Flexin's "Rubberz" hit #58 on the Billboard Hot 100, but faces accusations of being AI-generated, with experts citing unusual style, a fake British accent, and AI artifacts in the audio.

YouTuber Hank Green paused his channels, citing an unhealthy AI chatbot dependency after viewers flagged an AI-generated phrase in a July 29 video. Green admitted using ChatGPT for research, noting the "unhealthy" dopamine from LLM interactions.

OpenAI CEO Sam Altman's pitch for parents to use ChatGPT Work for children's podcasts backfired, as Alex Hirsch's viral reply questioning AI's role in parenting garnered significantly more engagement, revealing public skepticism about AI in family life.

Leopold Aschenbrenner's Situational Awareness hedge fund lost $35 billion in July due to an AI stock rout, leading to a fire sale of its public portfolio to Citadel. The fund's assets dropped from $45 billion to $10 billion.

LinkedIn has launched a new "Seems like AI slop" report button and updated its content policies to address the growing volume of AI-generated content on its platform. This move aims to improve content quality and authenticity for its professional user base.

FAR.AI's new AI Security Leaderboard, launched July 29, 2026, shows Grok and Gemini are vulnerable to automated jailbreaks, costing as little as $58 to exploit. In contrast, Claude Opus and GPT models resisted all attacks, highlighting significant differences in AI model security.

Apple CEO Tim Cook hinted at paid iCloud Plus tiers for heavy Apple Intelligence and Siri AI users during the Q3 FY2026 earnings call on July 30, 2026, as the company prepares for the iOS 27 launch this fall.

Scale AI appoints Francis deSouza, former Google Cloud COO, as its new CEO, effective August 10, 2026. This move signals a strategic pivot towards developing comprehensive AI applications for enterprises and governments.