AI News, November 8, 2026: ChatGPT Images 2.6, Gemini Omni Video and Claude Personal Finance

Short answer: The biggest AI stories for the week ending November 8, 2026 are OpenAI launching ChatGPT Images 2.6 with GIF creation on every plan, Google making Gemini Omni 1.2 Flash generally available for native video up to 15 seconds, Anthropic rolling out Claude Personal Finance in beta, OpenAI closing an $11 billion funding tranche at a $1.25 trillion valuation, and Google DeepMind reporting AI-found weaknesses in legacy 1024-bit RSA encryption. Details and sources for each story are below.

1. ChatGPT Images 2.6 Brings GIFs to Every Plan

OpenAI launched Images 2.6 on November 4, 2026, and it is the most useful release of the week for everyday users. Three things changed: ChatGPT can now turn images into GIFs, images generate 50% faster than before, and edits are steadier across revisions. The important part is the availability — OpenAI says Images 2.6 is live in ChatGPT across every subscription tier, including the free plan, not locked behind the expensive plan. The model is split in two under the hood: Flare, the fast one, and Sunburst, the careful one. One restriction to know: SVG export has real limits, so designers should check the export options before planning a workflow around it.

Why it matters: animated images and fast edits were a paid-tier perk; now every ChatGPT user gets them. If you make social content, product mockups or marketing visuals, this is the update to try first.

2. Gemini Omni 1.2 Flash: Google's Video Model Goes Generally Available

Google DeepMind made Gemini Omni 1.2 Flash generally available on November 5, 2026, on the Gemini API and Gemini for business on Google Cloud. It creates and edits video natively — up to 15 seconds long, with audio — from text, images, audio or video, just by chatting with it. Google says the price is about 30% lower than the previous Omni 1.1 model, and lists $0.07 per second for 720p video. Videos carry SynthID watermarks so AI-generated content can be identified.

Why it matters: video you can direct by conversation, at a lower price, inside Google's own stack. For marketers and creators, the 15-second limit fits short social clips — and the SynthID watermark answers the "is this real?" question before it is asked.

3. Claude Personal Finance Rolls Out in Beta — Read-Only

Anthropic began rolling out Claude Personal Finance on November 5, 2026, for US users on Pro, Max and Ultra plans. It connects to your bank accounts through Plaid and answers questions about your spending, subscriptions and budgets. Anthropic stresses two things: access is read-only — Claude cannot move money — and your financial data is not used for training. Plaid confirmed the partnership in a November 5 post.

Why it matters: it is the first time a major AI assistant has been given supervised, read-only access to real financial accounts. Useful for spotting subscriptions and spending patterns; still early, and still a beta.

Read Also: Best AI Tools for Small Business Owners in 2026  |  ChatGPT vs Claude: Which AI Is Better for Writing?

4. OpenAI Closes $11 Billion at a $1.25 Trillion Valuation

OpenAI closed an initial $11 billion tranche of new funding at a $1.25 trillion valuation, according to Bloomberg's November 6, 2026 report, with TechCrunch and The Verge separately reporting investor interest at the same price tag. The round makes OpenAI one of the most valuable private companies in the world and funds the compute build-out behind its models and products.

Why it matters: the valuation sets the scale of the AI race. Google answered the same week by buying a troubled Wyoming data center project from Crusoe — Reuters reported the acquisition on November 6, after Google announced a $17 billion investment in its first Wyoming data center. The infrastructure race is now as big a story as the models.

5. DeepMind's AI Finds Weaknesses in Legacy RSA-1024 Encryption

On November 6, 2026, Google DeepMind announced that human-guided AI agents had uncovered two major weaknesses in legacy 1024-bit RSA encryption, achieving the findings in just a few weeks. The important context: modern 2048-bit and larger keys remain secure and are not affected. DeepMind reported about a 1,600x theoretical improvement under short fingerprint control and roughly 30x lower marginal cost for batch factoring of additional 1024-bit keys.

Why it matters: it is a warning to anyone still running legacy 1024-bit keys, not a reason to panic about modern encryption. If your systems still use RSA-1024 anywhere, this is the week to retire it.

6. Perplexity Teammates and the Rest of the Week

Perplexity announced Teammates on November 4, 2026 — AI agents meant to own an ongoing job instead of answering one-off questions. Thinking Machines Lab shipped Inkling 2 Small on November 5: a mixture-of-experts model with about 594 billion total parameters and 17 billion active, with Apache 2.0 weights planned. OpenAI and Ligo published BioBasicsBench on November 6, a benchmark for routine wet-lab science tasks on which the best model scored only 66.7% — OpenAI itself says it is still too early to entrust experimental judgments to AI. Google also published a repository of Agent Skills for its products, and marked its Agent-to-Agent (A2A) protocol stable in ADK 2.14.0.

Security note: AppOmni found that Claude Code's "Don't Ask Again" feature led to more than 1,100 public GitHub files containing live passwords and API keys, many committed automatically by AI helpers. If you use AI coding agents, never let them commit credentials, and audit your repositories this week.

What This Week Means for You

For creators and small businesses, the practical takeaways are Images 2.6 and Gemini Omni 1.2 Flash: faster images, GIFs and conversational video editing are now available without enterprise budgets. For developers, the A2A protocol going stable and Claude Code 2.1.319 matter more. For everyone, the security story is the one to act on — AI agents are powerful enough to leak your secrets into public repositories if you give them broad permissions.

Related Guides

• ChatGPT vs Claude: Which AI Is Better for Writing in 2026?
• Best AI Tools for Small Business Owners in 2026
• Best AI Tools for Students in 2026

Frequently Asked Questions

What is ChatGPT Images 2.6?

Images 2.6 is OpenAI's image model launched on November 4, 2026. It generates images 50% faster, supports GIF creation and steadier edits, and is available on every ChatGPT plan including free. It comes in two variants, Flare (fast) and Sunburst (careful).

What is Gemini Omni 1.2 Flash?

Gemini Omni 1.2 Flash is Google DeepMind's model for creating and editing video by conversation. It became generally available on November 5, 2026, makes clips up to 15 seconds with audio, and costs about 30% less than the previous version — Google lists $0.07 per second for 720p video.

What is Claude Personal Finance?

A beta feature from Anthropic, rolling out from November 5, 2026, for US Pro, Max and Ultra users. It connects to bank accounts through Plaid and answers questions about spending and subscriptions. Access is read-only and the data is not used for training.

Is OpenAI really worth $1.25 trillion?

That is the valuation of its latest funding round, reported by Bloomberg on November 6, 2026, when OpenAI closed an initial $11 billion tranche. It reflects investor expectations about AI's future revenue, not current profit — treat it as a market signal, not a fact about the company's earnings.

Is my encryption still safe after the DeepMind RSA finding?

Yes, if you use modern standards. The weaknesses found apply to legacy 1024-bit RSA. The current standard, 2048-bit RSA and larger, is not affected. If any of your systems still use RSA-1024, retire it.

What happened with Claude Code and leaked secrets?

AppOmni found that Claude Code's "Don't Ask Again" feature led to more than 1,100 public GitHub files containing live passwords and API keys. The lesson: never let an AI coding agent commit credentials, and audit your repositories regularly.

All facts above were reported by the cited outlets (Bloomberg, TechCrunch, The Verge, Reuters, Google DeepMind, OpenAI, Anthropic, Plaid) in the week ending November 8, 2026. Vendor-reported figures are marked as such — check the original announcements before acting on them.