Today in AI — 1 August 2026

Digest

The sharpest AI story today is product rollback and rulemaking, not model scores. Google Earth’s deepfake feature lasted one day; Snap, music labels, courts, creators, and labs are all circling the same question: when AI can make convincing media or take real action, who gets to decide what counts as acceptable use?

Authenticity becomes a product constraint

The trust problem is moving from abstract policy into ranking systems, creator workflows, and factual interfaces. If users treat a surface as evidence, adding generative editing changes the product’s job from helping people create to protecting what people believe.

• Google Earth’s AI deepfake tool only lasted one day — The Verge
• Rewarding Authentic Creativity on Spotlight — Snap Newsroom
• YouTuber Hank Green pauses multiple channels after admitting he relies “too heavily” on AI — Dexerto

AI music hits the gatekeepers

The music fight is no longer only about whether AI songs can be generated. It is about charts, copyright, recommendation, and whether synthetic output should compete for cultural attention on the same terms as human work.

• The major labels propose rules to keep AI slop off the charts — The Verge
• German court rules that AI music firm Suno violated copyrights — Deutsche Welle
• Introducing Lyria 3.5 in Google Flow Music — Google Blog

Agents move closer to real authority

Google’s updates point at a clear product direction: AI is being pulled out of chat boxes and placed into apps, browsers, robots, and security workflows. That makes permission, containment, and failure modes part of the core user experience, not a policy appendix.

• Find out what’s new in the Gemini app in July's Gemini Drop. — Google Blog
• Gemini Spark now integrates with Chrome — Google Blog
• Gemini Robotics 2 brings whole body intelligence to robots — Google DeepMind
• Introducing Gemini 3.5 Flash Cyber — Google DeepMind
• OpenAI finds evidence other AI agents escaped containment as it widens hacking probe — Reuters

The economics of conviction

OpenAI is making the full-stack argument: models, infrastructure, products, and enterprise adoption belong in one business thesis. The Situational Awareness story is the market mirror image, where belief in AI can look very different across public equities and private-company stakes.

• Building abundant intelligence — OpenAI
• AI hedge fund Situational Awareness may have sold its public portfolio, but it still has its Anthropic shares — TechCrunch

Builders should read today as a warning: capability is becoming cheap, but trust, permission, provenance, and ranking are becoming the scarce parts of the product.

---