AI News Oct 28, 2026
Language: DE / EN / ZH / FR / ES / PT / JA / KO
AI-generated: summaries written by AI from the linked sources. How we use AI
Key Takeaways
- Google released Gemini 4 Carbon, a frontier model it says delivers significant gains in coding, agentic tasks, and cybersecurity. Google claims Carbon tops the Artificial Analysis …
- Artificial Analysis launched Intelligence Index v5, adding two new coding and science benchmarks and three upgrades, with most of the index now built on private datasets. Claude Op…
- Cybersecurity startup Cathedral used AI agents to breach Anthropic's internal systems through third-party software as part of a bug bounty program. Anthropic confirmed access to sa…
- Foundry: $100M (Series A)
- Marvell Technology → Lumentum Holdings
- When AI tools don't help a task, document the specific reasons and share them with peers and your manager instead of forcing adoption; explaining where AI falls…
Why Today Matters
- Benchmarks and model claimsGoogle's Gemini 4 Carbon tops the Artificial Analysis Intelligence Index v5.0, the company says, but the same index now leads with Claude Opus 5.5 at 58 and relies mostly on private datasets, so vendor benchmark claims deserve scrutiny.
- Agent security and exposureCathedral's AI agents reportedly reached Anthropic's sales, HR and IT data through third-party software, while Google now lets over 40 external agents into Android Studio, widening the same dependency surface.
- Enterprise AI commercializationAnthropic's new usage-visibility tools arrive ahead of a planned IPO, and Google is opening Gemini and Search to third-party apps, both moves aimed at making enterprise AI spending and distribution more legible.
- Open models and toolingNvidia released open-weight Nemotron 3.5 Transcribe and Granary v2, while GitHub made Copilot Fusion generally available with user-chosen effort levels, continuing the shift toward configurable, efficiency-focused model routing.
- AI regulation and safetyNew York's Hochul signed bills targeting AI chatbot features for minors and connected baby monitors, adding state-level rules as OpenAI plans an election war room for the US midterms.
What are the top AI breakthroughs?
This Oct 28, 2026 covers 10 curated AI news items spanning technology, research, and product developments. Google released Gemini 4 Carbon, a frontier model it says delivers significant gains in coding, agentic tasks, and cyber...
Google released Gemini 4 Carbon, a frontier model it says delivers significant gains in coding, agen…
Google released Gemini 4 Carbon, a frontier model it says delivers significant gains in coding, agentic tasks, and cybersecurity. Google claims Carbon tops the Artificial Analysis Intelligence Index v5.0, ahead of Claude Opus 5.5 and GPT-6 Astra.
Artificial Analysis launched Intelligence Index v5, adding two new coding and science benchmarks and…
Artificial Analysis launched Intelligence Index v5, adding two new coding and science benchmarks and three upgrades, with most of the index now built on private datasets. Claude Opus 5.5 (max) leads at a score of 58, according to the organization.
Cybersecurity startup Cathedral used AI agents to breach Anthropic's internal systems through third-…
Cybersecurity startup Cathedral used AI agents to breach Anthropic's internal systems through third-party software as part of a bug bounty program. Anthropic confirmed access to sales, HR, and IT data but disputes how deep the breach went, highlighting exposure through external software dependencies.
Google opened Android Studio to third-party AI agents via a 'Bring Your Own Agent' feature supportin…
Google opened Android Studio to third-party AI agents via a 'Bring Your Own Agent' feature supporting more than 40 coding agents, including Anthropic's Claude Agent and OpenAI's Codex. Google says new terminal tools help agents finish UI tasks 77% faster while using up to three times fewer tokens.
Nvidia released Nemotron 3.5 Transcribe, an open-weight model for multilingual audio transcription a…
Nvidia released Nemotron 3.5 Transcribe, an open-weight model for multilingual audio transcription and translation that the company says leads a benchmark leaderboard, along with Granary v2, its open-source speech dataset.
Anthropic launched tools that show companies how their spending on its models is being used, lifting…
Anthropic launched tools that show companies how their spending on its models is being used, lifting the curtain on usage ahead of a planned IPO. The move gives enterprise customers visibility into what their AI budgets are buying.
Google opened Gemini and Search to more apps through a new developer platform that lets businesses b…
Google opened Gemini and Search to more apps through a new developer platform that lets businesses bring their services into Google's AI experiences. The move could change how consumers discover and use apps as AI agents reshape the web.
GitHub made Copilot Fusion generally available, with the system handling model selection and orchest…
GitHub made Copilot Fusion generally available, with the system handling model selection and orchestration while developers choose the effort level, and released the GitHub Copilot app. An AT&T engineer cited improved token efficiency, according to the report.
New York Gov. Kathy Hochul signed two bills: one banning AI chatbots from using 'unsafe features' fo…
New York Gov. Kathy Hochul signed two bills: one banning AI chatbots from using 'unsafe features' for minors and another requiring security features in internet-connected baby monitors. The laws add to state-level efforts curbing AI risks for children.
OpenAI said it is launching an election 'war room' ahead of the US midterms to identify potential sp…
OpenAI said it is launching an election 'war room' ahead of the US midterms to identify potential spread of misinformation or other efforts to disrupt elections. The company disclosed the plan exclusively to Decoded by POLITICO.
Top AI Videos This Week
2 curated YouTube videos about AI developments.
Based on a transcript excerpt: Fireship breaks down three weeks of AI market turmoil, explaining the…
Based on a transcript excerpt: Fireship breaks down three weeks of AI market turmoil, explaining the circular financing arrangements behind GPU purchases and analyzing Anthropic's S-1 filing. Viewers learn how Nvidia's investments, GPU-backed borrowing, and Neo clouds create interconnected financial risk in the AI sector.
Based on a transcript excerpt: Matthew Berman demonstrates DeepSeek's latest open-source model, test…
Based on a transcript excerpt: Matthew Berman demonstrates DeepSeek's latest open-source model, testing its ability to generate a detailed interactive 3D mechanical watch diagram in a single web page. The video showcases the model's spatial awareness, 3D generation, and internal knowledge, positioning it as a leading open-source option.
What are the latest AI investment signals?
Latest AI investment signals: 2 funding rounds, 3 market updates, and 1 M&A transactions.
Primary Market – Funding Rounds
| Company | Amount | Round | Investors |
|---|---|---|---|
| Foundry | $100M | Series A | Marlinspike |
| Magic Vision | €600K | Unknown |
Secondary Market – Market Updates
M&A – Mergers & Acquisitions
| Acquirer | Target | Deal Value | Deal Type |
|---|---|---|---|
| Marvell Technology | Lumentum Holdings | Acquisition |
What are practical AI tips this week?
4 practical AI tips curated from Reddit communities and expert blogs. Terk Terkowitz of Amazon engineering leadership argues that if current AI tooling doesn't benefit yo...
Productivity
Terk Terkowitz of Amazon engineering leadership argues that if current AI tooling doesn't benefit your work, that's a finding rather than a failing, and engineers should document why and share it.
When AI tools don't help a task, document the specific reasons and share them with peers and your manager instead of forcing adoption; explaining where AI falls short demonstrates engineering judgment.
Creativity
Simon Willison reports that Claude Opus 5.5 shows meaningful improvements in 3D modeling, music generation, and game design, with game design being key to making games actually fun rather than just game-like.
When using coding agents to build games, focus the model's effort on the gameplay loop and design (what makes it fun) rather than only on visuals that merely look like a game.
Productivity
Oracle and OpenAI want employees to access business data and complete approved workflows through ChatGPT; individual Oracle plugins are listed for installation while the full AI Workspace experience remains under development.
Where available, install individual approved plugins to connect business data and run approved workflows through ChatGPT, rather than waiting for a complete integrated workspace that is still under development.
Productivity
Mithril Medicine had one full-time employee supported by outside specialists and laboratories; its acquisition by Anthropic is described as a case study in how Claude can help coordinate drug research, with clinical progress still to be tested.
Consider using Claude to coordinate research work across outside specialists and labs when a small core team needs to orchestrate distributed contributors, while treating clinical or outcome claims as untested.