🗞 Trump Scraps AI Controls, Nvidia's Saudi Deal, and Databricks' $1B+ Acquisition

🔎 The Latest on the AI Frontier:

• Trump Administration Scraps Biden-Era AI Export Controls

• Nvidia Secures Massive Blackwell Chip Deal with Saudi Arabia

• Databricks Expands AI Infrastructure with $1B+ Neon Acquisition

• Google Tests AI Mode Button to Replace "I'm Feeling Lucky"

• Musk's xAI Fails to Deliver Promised Safety Framework

• TikTok Introduces "AI Alive" Photo Animation Feature for Stories

• Other news you might find interesting

🔄 Trump administration officially ends Biden-era AI Diffusion rules that would have restricted US AI chip exports

• The US Commerce Department rescinded the controversial Framework for Artificial Intelligence Diffusion export controls that were set to take effect this week, stating they would have "stifled American innovation," created "burdensome regulatory requirements," and "undermined US diplomatic relations with dozens of countries."

• While the Trump administration has promised to issue replacement rules to safeguard national security interests, specific details remain limited, though the Commerce Department did issue guidance warning chipmakers they'll need permission before selling AI accelerators to foreign cloud providers if those chips might be used by hostile nations for training AI models.

• The Bureau of Industry and Security (BIS) also clarified that "in-country transfers" of existing GPU stockpiles by foreign cloud providers will require permission if those chips will be used to train "enemy AIs," and warned that using Huawei's Ascend AI accelerators anywhere in the world violates US export controls.

💻 Nvidia to sell 18,000 cutting-edge Blackwell AI chips to Saudi Arabia's Humain in major deal

• Nvidia CEO Jensen Huang announced the sale of over 18,000 of its latest GB300 Blackwell AI chips to Saudi company Humain during a White House-led trip to Saudi Arabia that included President Donald Trump and other top executives, with the chips to be used in data centers totaling 500 megawatts.

• The Saudi Public Investment Fund-owned Humain plans to develop AI models and build data center infrastructure with eventual plans to deploy "several hundred thousand" Nvidia GPUs, while AMD also announced it would supply chips to Humain as part of the $10 billion project.

• The announcement underscores Nvidia's chips becoming a powerful bargaining tool for the Trump administration as countries worldwide compete for the advanced AI hardware, with Nvidia shares rising more than 5% and AMD stock increasing 4% following the news.

🚀 Databricks acquires Neon, a serverless Postgres database provider, for over $1 billion

• Neon's technology can deploy a Postgres instance in under 500 milliseconds, making it ideal for AI agent workloads that require rapid database provisioning.

• The acquisition addresses traditional database limitations by separating compute from storage, allowing costs to remain proportional to actual usage while maintaining full compatibility with the Postgres ecosystem.

• The integration aims to help development teams build more efficient AI agent systems by preventing performance bottlenecks when thousands of agents run simultaneously, simplifying infrastructure, and reducing costs.

🔄 Google tests replacing its iconic "I'm Feeling Lucky" button with a new AI Mode button on its homepage

• The test includes two different designs: one that completely replaces the "I'm Feeling Lucky" button with an AI Mode button (featuring a colorful border), and another that adds an AI Mode shortcut to the search bar while keeping the traditional buttons underneath.

• Google confirmed to The Verge that this is "just one of many experiments" as they "often test different ways for people to access our helpful features."

• This potential change would represent a significant update to Google's famously minimal homepage that has remained relatively unchanged for decades, coming alongside other recent updates including Google's first "G" logo redesign in ten years.

🚨 xAI misses self-imposed deadline to release its final AI safety framework

• Elon Musk's AI company failed to deliver its promised AI safety report by May 10, drawing criticism from monitoring agency "Midas Project" as xAI continues to underperform in safety metrics compared to competitors.

• Despite Musk's frequent warnings about AI dangers, xAI's chatbot Grok has demonstrated poor safety practices, including inappropriate behavior when processing certain images and using crude language with frequent profanity.

• The trend extends beyond xAI, as competitors like Google and OpenAI have also shown signs of rushing safety testing and delaying model safety reports, raising expert concerns about potential risks as AI capabilities continue to advance.

🎬 TikTok launches AI Alive, a feature that transforms static photos into dynamic videos within TikTok Stories

• The new tool gives creators the ability to animate their photos with movement, atmospheric effects, and sound, allowing them to tell "richer, more visually engaging stories" - like turning a sunset photo into a cinematic clip with shifting colors, drifting clouds, and ambient wave sounds.

• Available exclusively in TikTok's Story Camera, the feature can be accessed by tapping the blue "+" on the Inbox or Profile page, selecting a photo from the Story Album, and using the AI Alive icon that appears in the right toolbar.

• TikTok emphasized its safety measures, including multiple trust and safety checks that review the uploaded photo, AI generation prompt, and resulting video before posting, as well as adding an AI-generated label and embedding C2PA metadata to ensure transparency even if content is shared off-platform.

More news you might find interesting:

• Nearly half of tech leaders have already begun deploying agentic AI, new EY survey finds.

• Perplexity quietly tests new "Projects" feature for building AI-powered apps and documents.

• Microsoft lays off over 6,000 employees (3% of workforce) to fund AI investments.

• Harvey AI adds Anthropic's Claude and Google's Gemini to its legal AI toolkit, expanding beyond OpenAI models.

• College students are using ChatGPT as their personal life coach, and it's not just for homework help.

• Notion launches new integrated AI toolkit featuring OpenAI's GPT-4.1 and Anthropic's Claude 3.7.

• Nine out of ten organizations plan to hire AI talent, with 60% already appointing Chief AI Officers.

• Zapier emerges as the essential integration tool for managing AI chaos and streamlining workflows.

• Tesla plans to resume shipping Chinese components for Cybercab and Semi production following US-China tariff truce.

• Agentic AI won't replace SaaS apps but will transform how we use them, experts say.

• Android's next big reinvention will focus on security and efficiency, not just AI hype.

Have any feedback? Send us an email