Breaking AI news, ChatGPT developments, machine learning breakthroughs, LLM advancements, and expert analysis on how AI is transforming technology and society.
Chinese AI startup Moonshot unveiled Kimi K3, a model with 2.8 trillion parameters that it says matches OpenAI and Anthropic. The announcement caused a market selloff, with chipmaker TSMC falling 7% and SoftBank 9%.
Meta Platforms Inc. is in early-stage talks with Anthropic to lease data centers in a deal potentially worth up to $10 billion over two years. Anthropic already has similar compute licensing deals with Google, SpaceX, Microsoft, and Amazon.
Economist Nouriel Roubini forecast that universal basic income or socialism will be necessary as AI revolutionizes work, with robots replacing a large chunk of the population in 20-25 years. He predicted GDP growth accelerating to 6% by 2040 and 10% by 2050.
At the Vatican-hosted Global Nobel Laureates Assembly, experts from OpenAI, Google DeepMind, and others signed the Rome Declaration for an Unarmed and Disarming Peace, addressing AI and nuclear war risks. The declaration builds on Pope Leo XIV's encyclical on protecting humans in the age of AI.
Anthropic announced revised Claude Fable 5 access plans, with Claude Max and Team Premium subscribers facing a 50% usage limit from July 20, while Pro and Team Standard subscribers lose access and receive a one-time $100 usage credit. The change marks a shift from all-you-can-use AI subscriptions to token-based pricing.
The five largest US Wall Street banks collectively reported $114 billion in capital markets revenue in the first half of 2026, up 31.5% year-over-year, driven by AI-related stock trading, dealmaking, and financing. Asia-focused AI investments are fueling record equity revenues for these banks.
Fraud expert David Maimon testified before the House Oversight Committee that the US federal government lacks policies to prevent AI-driven fraud against taxpayers. He described fraud as a 'durable, specialized criminal infrastructure' exploiting gaps between agencies. The hearing was part of the Trump administration's 'War on Fraud.'
Moonshot's Kimi AI model challenges the belief that the US leads China in AI development. At a Beijing event, Chinese AI leaders warned China still lags behind the US, with one executive saying the gap may widen. An Anthropic executive estimated Claude maker's technology is six to 12 months ahead of Chinese rivals.
Meta has patented AI technology that constantly analyzes a user's voice and surroundings to detect emotional state. The system records audible communications and combines them with contextual factors like time and location. Meta pitches it as a fitness device to tailor workouts.
Chinese President Xi Jinping gave his first speech at the World AI Conference in Shanghai, calling for open-source AI and global regulation. He criticized US AI protectionism and announced the launch of the World AI Cooperation Organisation with 29 member countries, pledging AI training for developing nations.