AI Weekly
AI News ·
This week in AI — Gemini Omni 1.2 Flash makes 15-second videos, OpenAI's Images 2.6 adds GIFs
Official announcements only, Oct 30–Nov 5, 2026. Google's cheaper, longer video model, OpenAI's Images 2.6, tighter ChatGPT safety, a shared robots.txt signal for AI training, and GPT-6 Cyber.
From official announcements, Oct 30–Nov 5, 2026: Google released Gemini Omni 1.2 Flash, which generates and edits videos up to 15 seconds and costs $0.07 per second of 720p video through the API, 30% less than Omni 1.1. OpenAI announced Images 2.6 with faster generation and GIF creation, expanded violence-prevention safeguards in ChatGPT, support for a shared robots.txt signal for AI training preferences, and GPT-6 Cyber, a model for finding and fixing vulnerabilities.
We only cover what companies published on their own blogs and newsrooms, and every item links to the original. Quality claims below are the companies' own, not independent tests.
OpenAI's article pages blocked automated access this week, so for OpenAI items we only use the summaries in its official RSS feed — nothing beyond that.
1. Google's Gemini Omni 1.2 Flash: 15-second videos, 30% cheaper (Nov 5)
Google released Gemini Omni 1.2 Flash, a video model that generates and edits clips up to 15 seconds, up from 10 seconds in Omni 1.1.
- Length: pick any duration from 3 to 15 seconds, in one-second steps.
- Edit by chatting: refine a clip over several turns — swap a product or background, change the lighting or weather, switch the aspect ratio.
- References: guide it with character, product or style images, short motion clips (up to three seconds) and audio cues.
- The claims: more natural dialogue, clearer lip-sync, and far fewer unwanted jump cuts during camera moves. Google shows these with example clips, not measured scores.
- Where: the Gemini app (rolling out to Google AI Ultra subscribers), Google Vids (15-second clips at no extra cost with a Google account or Workspace), Flow, Google Ads, the Gemini API and Google AI Studio, and Google Cloud.
- API price: $0.07 per second of 720p video, 30% less than Omni 1.1.
- Outputs carry an invisible SynthID watermark. Google says fast sports action and complex physics are still being improved.
What it means for you: Google Vids is the no-cost way in. If you need a short product or explainer clip, try it there before paying for anything.
Source: Gemini Omni 1.2 Flash — Google
2. OpenAI's Images 2.6 adds GIFs and sharper edits (Nov 4)
According to OpenAI's summary, Images 2.6 brings faster image generation, GIF creation, more precise editing and new API controls, across ChatGPT, Codex and the API. The summary doesn't say which plans get it.
What it means for you: If you already make images in ChatGPT, try a GIF or a small targeted edit — those are the new parts.
Source: Introducing Images 2.6 — OpenAI
3. ChatGPT gets stronger violence-prevention safeguards (Oct 30)
OpenAI says it is expanding safeguards against violence in ChatGPT, intervening earlier, and committing to public safety reporting. The post is titled "Community safety report and new protections for teens," but the summary doesn't spell out the teen protections.
What it means for you: If your kids use ChatGPT, read the original for the teen details — the summary alone doesn't cover them.
Source: Community safety report and new protections for teens — OpenAI
4. One robots.txt signal to opt out of AI training (Oct 30)
OpenAI will honor a shared robots.txt signal that lets website owners state their AI training preferences to every participating provider at once.
What it means for you: Run a blog or a store? You get one more way to say whether your pages can be used for AI training. Check the original for which providers take part.
Source: Making it easier for websites to express AI training preferences — OpenAI
5. GPT-6 Cyber, a model for security teams (Nov 5)
OpenAI introduced GPT-6 Cyber, a cybersecurity model that finds exploitable vulnerabilities and develops fixes, trained to stay within authorized scope. Separately, OpenAI is raising its bug bounty rewards and adding bonuses for report quality, loyalty and priority campaigns.
What it means for you: Not a model you'll use directly. It does mean AI is now part of the race to find security holes before attackers do.
Source: GPT‑6 Cyber — OpenAI
The takeaway
AI can now make 15-second videos and GIFs, and the same week brought a lot of guardrails: safety reporting, training opt-outs and AI-powered security. New features only help if you use them well — find out your AI type in 3 minutes.
Sources
- Gemini Omni 1.2 Flash: Fast and cost-effective with longer clips — Google (2026-11-05)
- Introducing Images 2.6 — OpenAI (2026-11-04)
- Community safety report and new protections for teens — OpenAI (2026-10-30)
- Making it easier for websites to express AI training preferences — OpenAI (2026-10-30)
- GPT‑6 Cyber: Advancing the frontier of aligned cybersecurity agents — OpenAI (2026-11-05)
- Rewarding defenders with bigger bug bounties — OpenAI (2026-11-02)
- OpenAI News RSS feed — OpenAI
Written with AI (Claude) and checked against the linked sources.