Daily AI brief

AI advertising expands, security scanning sheds a prerequisite

OpenAI outlines advertising products, GitHub broadens AI Scan eligibility, Google introduces dialogue models, and Jensen Huang argues for vendor-led safety.

2026-09-16·Updated 2026-09-16·Primary sources shown for every item

01Products
OpenAI introduces Sponsored Agents and advertising integrations
OpenAI introduced AI advertising experiences spanning Sponsored Agents, marketer tools and integrations with HubSpot and Shopify. Its announcement identifies the product direction without specifying rollout coverage or how the sponsored experiences work.
Why it mattersFor marketing teams, the named integrations make compatibility with existing workflows a concrete evaluation question. Whether these products warrant adoption depends on what the connections actually let teams do.
Read analysis
SourcesOpenAI (opens in a new tab)
02Products
GitHub removes the CodeQL default-setup requirement for AI Scan
GitHub says AI Scan can now check pull requests for security vulnerabilities even when a repository has not enabled CodeQL default setup. The change removes that prerequisite; the notice does not establish scanning accuracy or describe other eligibility requirements.
Why it mattersRepository maintainers can evaluate AI Scan without first enabling CodeQL default setup. That changes the setup decision for teams seeking pull-request security checks, while detection quality remains a separate question.
Read analysis
SourcesGitHub Changelog (opens in a new tab)
03Models
Google introduces Gemini 3.8 Live and Live Extended Thinking
Google DeepMind introduced Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking as models built for natural conversation. Google calls them its most advanced live dialogue models, but the announcement gives no performance measurements or access details.
Why it mattersTeams building conversational interfaces have two named models to investigate. A useful selection decision will require evidence about their differences, response times and access conditions rather than relying on the Extended Thinking name.
Read analysis
Related directory entriesGemini — Google's assistant, tightly integrated with Search, Gmail and Workspace.
SourcesGoogle DeepMind (opens in a new tab)
04Policy & safety
Jensen Huang argues AI makers can handle safety without regulation
Nvidia's Jensen Huang argued against AI regulation, saying product makers can engineer safety themselves, TechCrunch reports. He framed AI as hardware and software rather than an alien form of intelligence, making a case for vendor responsibility rather than announcing a policy change.
Why it mattersThe position puts vendors' engineering practices at the center of the safety debate. Buyers assessing that approach still need concrete evidence of safeguards and accountability before treating a supplier's assurances as sufficient.
Read analysis
SourcesTechCrunch (opens in a new tab)

OpenAI introduces Sponsored Agents and advertising integrations

GitHub removes the CodeQL default-setup requirement for AI Scan

Google introduces Gemini 3.8 Live and Live Extended Thinking

Jensen Huang argues AI makers can handle safety without regulation