Submit a toolJoin free

Today in AI

Google removed an AI image feature from Google Earth one day after launch while it strengthens safeguards. Major record labels proposed keeping heavily AI-generated music out of mainstream charts. The EU AI Office gained powers to test advanced models, demand access and restrict risky deployments. DeepSeek’s low-cost V4 Flash coding model increased pressure on premium AI pricing. ChatGPT gained browser-context features that move it closer to becoming an intelligent browsing layer.

The latest signal

Top AI News

Five news stories from today’s digest.

AIMedia

Google removes Earth AI image feature after one day

TechCrunch•31 Jul 2026

News Digest

Google removed a Google Earth feature that generated synthetic imagery over real maps only one day after launch. The rollback followed criticism and problematic screenshots, while Google develops stronger safeguards. Because Earth is widely treated as a factual geographic reference, generated images could be mistaken for evidence. The tool remains unavailable during the review.

Key takeaways

03
  1. Google withdrew the Earth feature one day after launch because it placed synthetic imagery over maps commonly treated as factual references.
  2. Problematic screenshots circulated after release, prompting Google to develop stronger safeguards before considering whether the tool should return.
  3. The reversal shows that generative features carry greater risk when their output appears inside products users treat as evidence of the real world.

Why it matters

The problem is not image generation alone, but where its output appears. Synthetic content inside a trusted reference product can look like evidence, strengthening the case for visible labels, provenance controls and tighter limits when users may confuse generated material with the real world.

Read original story
AIMedia

Record labels propose limits for AI music on charts

The Verge•1 Aug 2026

News Digest

Major record labels proposed rules that would exclude most heavily AI-generated music from mainstream charts without banning its distribution. The plan aims to keep rankings focused on human-led popularity as automated production becomes easier to scale. It would also force platforms and rights bodies to define what meaningful human authorship requires.

Key takeaways

03
  1. Major labels want heavily AI-generated music excluded from mainstream charts, while allowing the same tracks to remain available for distribution.
  2. The proposal seeks to protect rankings as measures of human-led popularity while automated music production becomes easier to scale widely.
  3. Clearer authorship rules could influence discovery, royalties and commercial value by separating human-led releases from largely automated work.

Why it matters

Chart rules influence recognition, discovery, royalties and the commercial value attached to a release. Clearer authorship standards would help cultural markets distinguish between human-led music and work produced mainly through automation.

Read original story
AIRegulation

EU gains new enforcement powers over advanced AI

EnterpriseAI.com•1 Aug 2026

News Digest

From 2 August, the EU AI Office can test advanced general-purpose models, request technical access from providers and restrict deployment where risks remain unresolved. The European Commission can require corrective action and issue fines for prohibited practices or failure to cooperate. Certain generated content, including deepfakes, must also be clearly labelled.

Key takeaways

03
  1. From 2 August 2026, the European Commission can enforce obligations covering providers of powerful general-purpose artificial intelligence models.
  2. The AI Office may request technical documents, evaluate models, require corrective action and issue fines when providers fail to cooperate.
  3. Transparency rules require certain generated material, including deepfakes, to be marked clearly so people can identify synthetic content.

Why it matters

The AI Act is moving from written obligations to active enforcement. Model providers operating in Europe need stronger documentation, risk controls and transparency processes, while businesses using their systems may face closer scrutiny over how generated content is identified.

Read original story
AIModels

DeepSeek’s bargain coding model escalates AI price war

Axios•1 Aug 2026

News Digest

DeepSeek’s official V4 Flash model entered public beta on 31 July, replacing the preview version with stronger coding and agent capabilities. It costs about $0.28 per million output tokens, versus the $25 Axios cited for Anthropic’s Claude Opus 4.8. The model offers a one-million-token context window, tool calls and reasoning or non-reasoning modes.

Key takeaways

03
  1. DeepSeek’s official V4 Flash API entered public beta on 31 July, replacing the preview release with stronger coding and agent capabilities.
  2. Pricing starts at $0.28 per million output tokens and $0.14 per million uncached input tokens, far below the $25 cited for Claude Opus 4.8.
  3. Axios says falling model prices pressure premium providers to justify higher fees through greater accuracy, safety or operational reliability.

Why it matters

Lower inference prices make advanced coding models accessible to more businesses and developers. They may also encourage teams to route different tasks between several models, choosing according to cost, performance and risk rather than relying on one provider.

Read original story
AIAgents

ChatGPT is becoming an agentic browser

ChatGPT X•1 Aug 2026

News Digest

OpenAI is giving ChatGPT greater awareness of browser context, allowing it to reference open tabs, highlighted webpage text and YouTube videos. The desktop app can also suggest URLs, reducing the need to move repeatedly between tabs and search results. Together, the features position ChatGPT as an intelligent browsing layer rather than a conventional chatbot.

Key takeaways

03
  1. ChatGPT can use browser context, including open tabs and highlighted webpage text, to answer questions with more relevant information than before.
  2. YouTube understanding and desktop URL suggestions reduce the need to switch repeatedly between browser tabs, search results and copied text.
  3. OpenAI is positioning ChatGPT as an AI layer for navigating, understanding and interacting with the web, rather than only answering questions.

Why it matters

AI is moving closer to becoming the main interface for browsing the internet. As assistants interpret tabs, videos and highlighted text, users may rely less on manual searches and copying information between pages, shifting how people discover and navigate online content.

Read original story