🎉 Meta’s AI Ad Factory, Bing Launches Free AI Video Tool, OpenAI Adds Cloud & Meeting Tools, Google Upgrades Gemini, Claude Goes to Washington
Meta’s AI Ad Factory, Bing Debuts AI Video, OpenAI Expands ChatGPT Tools, Google Updates Gemini, Anthropic Launches AI for Gov
Welcome to this week’s edition of AImpulse, a five point summary of the most significant advancements in the world of Artificial Intelligence.
Here’s the pulse on this week’s top stories:
What’s Happening: Meta plans to launch AI tools by 2026 that remove humans from the ad creation process, according to WSJ — aiming for an AI that builds Facebook and Instagram ads using only a product image and a budget.
The details:
Companies will upload a product image and budget, and the AI will generate text, visuals, audience targeting, and manage placements.
Ads will personalize in real-time, tailoring content like showing a car on mountain roads or city streets based on user location.
The service is geared toward small businesses lacking marketing teams, offering professional-grade ads without agency costs.
Advertising already accounts for 97% of Meta’s revenue and remains central to Zuckerberg’s AI roadmap.
Why it matters: AI has already reshaped ad creation with image, video, and text generation — but Meta’s plan fully automates the process. Given the scale of ad spend on Facebook and Instagram, a working system could upend the market, especially for small brands that need fast, effective ads without extra overhead.
What’s Happening: Microsoft has rolled out Bing Video Creator, adding OpenAI’s Sora model into its mobile Bing app — letting users generate 5-second videos from text prompts, no subscription needed.
The details:
Users get 10 quick generations and unlimited slower ones, earning more fast credits via Microsoft Rewards.
The tool debuts on Bing’s iOS and Android apps, with desktop and Copilot Search integration on the way.
Videos are vertical-only for now, capped at 5 seconds, and users can generate up to three clips at once.
Why it matters: Despite early hype, Sora was quickly overtaken by rival models. However, most video generators sit behind paywalls — meaning this free, if limited, feature could bring AI video creation to a broader audience for the first time.
What’s Happening: OpenAI announced a wave of new business features for ChatGPT, including integrations with Google Drive, Dropbox, and other cloud services, plus meeting recording and custom enterprise connectors.
The details:
ChatGPT now connects directly to cloud services like Drive, Dropbox, and SharePoint for seamless document integration.
A new recording mode captures meetings or voice notes, generating transcripts, summaries, and action items.
Deep Research connectors are available for platforms like Outlook and Teams, with customizable MCP connectors for admins.
OpenAI also disclosed it now serves 3 million paying business users across its Enterprise, Team, and Education plans — up from 2 million in February.
Why it matters: OpenAI is streamlining data access and document management, reducing friction with new native integrations. The move also consolidates functions typically handled by multiple SaaS tools, positioning ChatGPT as a centralized enterprise workspace.
What’s Happening: Google previewed an update to its Gemini 2.5 Pro model, calling it its “most intelligent yet” — with major improvements in coding, STEM tasks, reasoning, and image comprehension.
The details:
The update brings strong gains across leaderboards like LMArena and WebDevArena.
Google addressed feedback from earlier versions to fix regressions in creative writing and non-coding tasks.
A new “thinking budget” feature lets developers control cost and latency through the API.
The preview is available now through the Gemini API on AI Studio and Vertex AI, and rolling out to the Gemini app.
Why it matters: Weeks after releasing 2.5 Pro’s I/O update, Google is back with broader improvements. Frequent preview updates signal a shift in model release strategies — aiming for continuous performance gains rather than infrequent big launches.
What’s Happening: Anthropic introduced Claude Gov, a special version of its AI model built specifically for U.S. defense and intelligence agencies — with customized safety guardrails and enhanced capabilities for handling classified data.
The details:
Claude Gov is already deployed at top U.S. national security levels for classified material.
The model offers lower refusal rates and better comprehension of defense and intelligence documents.
Key upgrades support foreign language analysis and cybersecurity pattern detection for intelligence work.
Anthropic carved out specific allowances for government contracts, while maintaining bans on weapons design, disinformation, and malicious cyber activities.
Why it matters: As AI firms tailor models for military and intelligence use, the technology is embedding deeper into government operations. But with it comes a balancing act — navigating the tension between commercial expansion and ethical boundaries.