Grok Bots Get Direct Access to Microsoft Accounts

Advertise here | 6-min Read

AI answers read a brand as one presence, not one channel at a time. Bring all your data together.

Run social, search, and AI visibility as one story. Connect Planable and SE Ranking to Claude or ChatGPT, and the research, the cross-channel patterns, and the drafts all happen in one conversation. Planable holds social performance. SE Ranking holds search demand and AI visibility. Bring all data together.

Which topic is Reddit carrying that we’ve never ranked for? Where did a post move before the keyword data caught up? Which post actually sent people to pricing? Find your answers in one chat by connecting Planable and SE Ranking MCPs.

Grok Bots Get Direct Access to Microsoft Accounts

AI CHEAT SHEET

Google Gemini Omni 1.1 Flash goes GA — 40-second scene extension,

Google released Gemini Omni 1.1 Flash on August 27 as the generally available production model for AI video generation and editing. The release moves Omni from a capable generator to a directable one: scene extension now reads up to 10 seconds of prior context instead of a single final frame, first and last frames can be pinned to control camera movement, drafts render in 360p at a third of 720p cost, finals upscale to 4K, and video clips can be passed as references for character consistency. The stable API identifier is gemini-omni-1.1-flash. The preview endpoint deprecates September 30.

The details:

• Gemini Omni 1.1 Flash can now extend a scene by up to 40 seconds, use keyframes to control transitions between shots, create low-cost 360p drafts for faster iteration, and upscale finished work to 4K. The 40-second extension is a significant jump previous Veo models used only 1 second of context; Omni 1.1 uses 10. The difference is visible in continuity

• First and last frame pinning lets developers specify exactly where a shot starts and ends, generating the motion in between rather than asking the model to invent the entire shot from scratch. This enables camera movements, zooms, and transitions with predictable endpoints — the single most-requested professional video control that previous AI video models could not reliably deliver.

Why it matters: AI video has been winning on impressiveness and losing on controllability. Omni 1.1 is the first production-ready model from a major lab that addresses the control problem directly pinnable frames, selectable resolution, character reference clips, 10-second context extension.

AI NEWS HIGHLIGHT

• OpenAI is ending its Cursor partnership — November 12 cutoff, SpaceX acquisition violated terms of service — Cursor built on OpenAI models. SpaceX now owns Cursor. OpenAI says that crossed a compliance line. Developers have 10 weeks to transition away from Cursor's OpenAI-powered features.
• NVIDIA reportedly acquiring Hugging Face for $12.9B — open-weight infrastructure is the hottest M&A target in AI — Stripe bought OpenRouter for $7B. NVIDIA buying Hugging Face for $12.9B. The platform that hosts open-weight models is now worth more than most of the labs that trained them.
• Anthropic used Claude as an automated alignment researcher — improved all 10 safety benchmarks without reducing capabilities — Claude's best proposals outperformed experienced human safety researchers. AI improving its own safety training is the most significant interpretability result since the "thinking room" discovery.
• Anthropic overtook OpenAI in business subscriptions in May — tracking $47B ARR, profitable 2029 — ChatGPT monthly visits below majority of market for the first time. Pentagon excluded Anthropic. Enterprise chose it anyway.
• Gemini Omni 1.1 Flash GA — 40-second scene extension, first/last frame control, 4K upscaling — preview endpoint deprecated September 30. Adobe, Figma Weave, Runway already in production. Draft at 360p, render at 4K.

Evaluate your AI vendor safety posture

Prompt: The Pentagon launched GenAI.mil today with ChatGPT Mil, Grok for Government, and Google Gemini — and excluded Anthropic's Claude because of its responsible scaling policies. A federal judge ruled the exclusion illegal yesterday. The contract launched today without Claude anyway. The FLI Safety Index in August gave no lab above C+. Claude Code Opus 5 Auto Mode has a published 60–80% exploit chain. The Summer 2026 Safety Crisis documented six incidents across every major lab. Anthropic overtook OpenAI in business subscriptions in May.

We are evaluating AI vendors for: [describe your use case — enterprise deployment, government contract, customer-facing product, or internal tool — and your sensitivity requirements].

Help me build a procurement framework across three dimensions:

1. Safety posture vs capability trade-off — The GenAI.mil decision chose capability and access over safety constraints. For our use case: what is the correct weighting? For consumer-facing AI in regulated industries, a vendor's FLI Safety Index score and incident history are material procurement facts. For internal developer tools, the published exploit rate of the tool matters more than the lab's governance score. Map our use case to the right evaluation criteria — not the criteria that apply to the DoD, and not the criteria that apply to a consumer health app, unless one of those is actually us.

2. The exploit surface audit — Claude Code Auto Mode has a published 60–80% arbitrary code execution exploit. GitHub Copilot Autofix introduced a shell-injection vulnerability. CoSnitch auto-exfiltrated connected accounts. For each AI coding or automation tool our team uses: what is the known exploit surface, and what configuration change reduces our exposure? "We trust the vendor to patch it" is not a configuration. It is hope.

3. The September routing decision — Sonnet 5 is now $3/$15. Claude is absent from GenAI.mil but leads in enterprise subscriptions. Gemini is in GenAI.mil but Gemini 3.5 Pro has missed six delivery dates. GPT-5.6 Sol is in GenAI.mil and is the fastest frontier model on Cerebras. For our specific workloads: which vendor's September 1 position — capability, pricing, safety, and procurement signals — makes them the right choice for the next six months?

End with a one-paragraph vendor selection rationale — specific enough that a procurement officer could sign it, honest enough that a safety auditor could review it.

TOP TRENDING AI TOOLS

Research & Content

•Clipto— Local AI search for video, audio, meetings, files — on-device, no cloud
•Claude— Overtook OpenAI in business subscriptions in May — enable hardware key, check Claude Code Auto Mode settings

Software Development

•Cursor— — Origin opt-out recommended until data terms published
•Construct Computer— Multi-agent automation, scheduled jobs, persistent context
•Murmell— Cloud canvas for team and AI agents on shared codebases

Agents & Automation

•AirJelly— Proactive second brain — on-device memory, screen-aware, local execution
•Project SKY— Ambient AI companion for Windows — on-device memory, proactive tasks
•Wispr Flow— Dictate in any app — 4x faster than typing, 100+ languages

Video & Productivity

•Soloop— AI video editor — auto-cuts, captions, repurposes long-form content
•ElevenLabs— Industry standard for voice cloning and text-to-speech

That’s a Wrap

SPONSOR US

Get your business in front of over 90k+ AI professionals

8020AI is the world’s #1 AI Newsletter, Read by 90k+ professionals from leading companies such as Google, OpenAI, Meta, and Microsoft.

We've assisted in promoting Over 500 AI-Related Products. Will yours be the next?

What We Can Offer:

• Launch an Advertising Campaign

• Introduce New Product or Features

• Other Business Cooperation

Get in touch today.

Or Email our founder Alamin at [email protected]

FEEDBACK

How was your experience with 8020AI today?

Login or Subscribe to participate in polls.

If you have specific feedback or anything interesting you’d like to share, please let us know by replying to this email.