Weekly Tech & AI Newsfeed

AI in Entertainment: Google DeepMind and A24 Announce $75M Filmmaking Partnership

Artificial intelligence is rapidly expanding into the entertainment industry as Google DeepMind partners with independent film studio A24 to develop advanced AI filmmaking tools. This multiyear research collaboration reportedly includes a $75 million investment from Google to help artists create next-generation creative workflows.

As part of the initiative, A24 is prototyping an AI storyboard application designed to identify logistical production challenges before shooting begins. To proactively address AI copyright concerns, the studio’s existing film library will remain strictly off-limits for Google’s machine learning training models. Ultimately, this strategic partnership ensures that future generative AI technologies are shaped directly by the filmmakers who use them.

Sources:

AI Cybersecurity: Agent CI Bench Exposes Privacy Data Leaks in Frontier AI Agents

A groundbreaking evaluation framework called Agent CI Bench has revealed significant AI privacy risks, finding that 12 out of 15 frontier computer-use AI agents leak private data in over 50% of tested privacy scenarios. The benchmark tested AI agents across a controlled workspace of applications and uncovered an alarming average data leakage rate of 67.9%.

Researchers noted a concerning disconnect between an AI model’s task completion capabilities and its data security:

  • Claude-Opus-4.7 demonstrated a 14.0% leakage rate with 87.4% utility.
  • Gemini-3.1-Pro achieved 96.6% utility but suffered a massive 98.3% leakage rate.

The study also highlighted that simple prompt-level mitigations can reduce these AI privacy leaks by 33 to 36 percentage points while simultaneously improving overall task utility.

Sources:

Cloud Infrastructure: SpaceX Secures $6.3 Billion AI Compute Deal with Reflection AI

SpaceX is aggressively expanding its footprint in the AI cloud infrastructure market, signing a massive $6.3 billion computing agreement with open-source AI startup Reflection AI. Starting July 1, 2026, Reflection will pay $150 million monthly to access high-performance Nvidia GB300 chips at SpaceX’s Colossus 2 data center in Memphis, Tennessee.

The AI compute contract, which runs through 2029, allows either party to terminate with a 90-day notice after the first three months. This landmark agreement further cements SpaceX’s position as a dominant player in AI data centers, adding to its existing multi-billion dollar cloud compute deals with tech giants like Google and Anthropic.

Sources:

US Technology Policy: Landmark Quantum Computing Executive Order Signed

On June 22, 2026, U.S. President Donald Trump signed a pivotal executive order focused on advancing quantum computing research, emerging technologies, and national cybersecurity. The directive establishes the Quantum Computer for Application Development and Discovery Science (QC-ADDS) Effort, which aims to engineer the first quantum computer powerful enough for advanced scientific research by 2028.

Furthermore, the executive order seeks to bolster quantum cybersecurity to protect national defenses against future quantum systems capable of breaking current encryption standards. Attended by key officials, including National Cyber Director Sean Cairncross and Secretary of Commerce Howard Lutnick, the initiative is designed to strengthen U.S. competitiveness and technological leadership on a global scale.

Sources:

Open-Source AI: GitHub Coalition Pushes to Amend California AI Transparency Act

GitHub has partnered with Black Forest Labs, Hugging Face, and Mozilla to form a coalition advocating for targeted amendments to the California AI Transparency Act. The open-source group argues that the proposed AI legislation’s license revocation provisions conflict with standard open-source licensing norms, which are inherently perpetual and irrevocable.

If passed without changes, the bill could introduce significant uncertainty across the global software supply chain and disrupt community-driven open-source AI projects. To better protect the open-source ecosystem while maintaining regulatory intent, the coalition proposes aligning the California act with the European Union’s AI Act Transparency Code of Practice.

Sources: