Welcome, AI Visionaries!

The future of AI is being written in Washington, Silicon Valley, and beyond. This week, AI leaders made bold moves that could shift the global balance. From intense calls to fast-track infrastructure investment to game-changing technology designed to reshape how AI models evolve, it’s clear that the race for AI dominance is heating up.

In today’s Generative AI Newsletter:

• AI CEOs push Congress to loosen rules or risk losing the global race
• OpenAI hires Instacart’s CEO to lead consumer product strategy
• Alibaba’s ZeroSearch redefines how AI models train themselves
• Microsoft bans DeepSeek due to security concerns amid AI data scrutiny

Special highlight from out Network

Proven RAG at scale. Trusted by Perplexity

Scaling RAG from pilot to production exposes challenges:

latency, cost, accuracy, and adaptability.

Perplexity tackles these at scale, processing 100M+ weekly queries from 15M+ users—powered by Vespa.ai’s real-time platform.

Proven in production, Vespa delivers:

• Fast, accurate retrieval
  

• Scalable infrastructure for generative AI, copilots, assistants, and search
  

The result: teams move confidently from prototype to full deployment.

Ready to scale RAG like a pro?

🏛️ AI CEOs Warn Congress: Loosen AI Rules or Lose to China

Some of the biggest names in AI from OpenAI, Microsoft, AMD, and CoreWeave showed up on Capitol Hill this week, urging lawmakers to move faster on AI infrastructure, loosen export rules, and cut red tape. The message was clear: if the U.S. wants to stay ahead of China, it can’t afford to slow AI down.

What they said:

• Sam Altman (OpenAI) called AI “bigger than the internet” and asked Congress to invest in the foundations that will support it, such as chips, power, and talent.

• Brad Smith (Microsoft) warned that current chip export bans are pushing global customers toward Chinese alternatives, which is exactly the opposite of what the U.S. wants.

• Lisa Su (AMD) backed him up, saying if U.S. tech gets blocked, the world will just turn elsewhere.

• All four pushed for bigger investments in AI research, a stronger electric grid, better permitting processes, and training programs to grow the AI workforce.

There’s a real fork in the road here. Regulators want to play it safe with guardrails and oversight. But the industry is saying go too slow and China wins. Congress now has to decide whether to empower these companies or try to rein them in, knowing either path comes with real consequences.

🛒 OpenAI Taps Instacart CEO to Lead Its Consumer Strategy

OpenAI just made a major leadership move by hiring Instacart CEO Fidji Simo as its first-ever CEO of Applications. It is a big step that signals how serious the company is about turning its research breakthroughs into global consumer products.

What to know:

• Simo will run a new division that brings together OpenAI's business and product teams under one roof. Her focus will be taking AI from the lab to the everyday user.

• She will report directly to Sam Altman, who said this shift lets him spend more time on the technical side of things, including research, infrastructure, and safety.

• Simo is no stranger to OpenAI. She has served on its nonprofit board for a year and previously spent a decade at Facebook leading its core app and ads business.

• The hire follows OpenAI’s decision to slow its for-profit push and ramps up just as its global Stargate infrastructure project kicks into gear.

OpenAI is evolving into a mature company with real organizational depth. Simo brings the kind of operational muscle the company needs if it wants to reach billions of users. Altman may still be the visionary, but OpenAI is building the kind of leadership bench you need when you are playing to win on a global scale

🔎 Alibaba's ZeroSearch Trains AI Without the Internet

Alibaba has developed a new method called ZeroSearch that lets AI models teach themselves how to search for information without ever touching a real search engine. It is a major cost saver and could reshape how future AI tools learn.

What to know:

• ZeroSearch replaces costly search engine API calls with simulated results generated by another large language model.

• It uses a clever training approach that gradually feeds the AI lower quality documents, forcing it to sharpen its reasoning skills over time.

• The result is a huge cost reduction and tighter control over the training process, avoiding the noise and inconsistency of real-world search data.

• In tests, the models trained with ZeroSearch performed as well or better than those trained with actual search engines, while slashing costs by 88 percent.

This is one of those quiet breakthroughs that could level the playing field. Training AI to search without relying on Google or Bing makes advanced systems more affordable and accessible. If this method scales, the next generation of AI may learn more from simulations than the messy real world.

🚫Microsoft Employees Are Banned from Using DeepSeek

Microsoft is taking a firm stance against DeepSeek, the controversial app, with its president, Brad Smith, announcing in a Senate hearing that employees are not allowed to use it due to concerns over data security and potential propaganda.

What to know:

• Brad Smith stated that Microsoft has refrained from including DeepSeek in its app store, citing the risk of user data being stored in China and influenced by government-backed narratives.

• DeepSeek’s privacy policy reveals that it stores data on Chinese servers, where it is subject to Chinese law, including requirements for cooperation with the country’s intelligence agencies.

• Despite these concerns, Microsoft made DeepSeek’s R1 model available on its Azure cloud platform earlier this year, though offering the model on a cloud service is a different scenario than supporting the app itself.

• Smith also noted that Microsoft had made adjustments to DeepSeek’s AI model to eliminate “harmful side effects,” although he did not provide further specifics.

Microsoft’s decision is a reminder of how AI models, even when offered on a cloud service, carry broader risks. Microsoft’s selective restrictions raise questions about the balance between competitive interests and safeguarding user trust, especially considering the company's own AI initiatives.

🚀 Boost your business with us—advertise where 10M+ AI leaders engage

🌟 Sign up for the first AI Hub in the world.

📲 Our Socials