Welcome, AI Enthusiasts!
A $200M DoD contract has pulled OpenAI straight into the machinery of military AI. Google’s Gemini 2.5 stack is now fully live, with a new ultra-fast Flash-Lite aimed at volume. The new American Pope is making AI a central issue of faith and policy. And Search itself is talking back.
In today’s Generative AI Newsletter:
OpenAI inks $200M Pentagon deal under new ‘Gov’ division
Gemini 2.5 stabilizes, launches Flash-Lite for cheap speed
Pope Leo XIV confronts AI’s moral threat at Vatican summit
Google Search goes live with AI-powered voice chat
Special highlight from our network
LeadWithAIAgents Hackathon – Global + Hybrid
Your launchpad to build the next wave of AI tools.
This one’s for the doers. Over 72 hours, you’ll join top engineers, PMs, and product teams to build AI agent-powered projects with real-world potential.
Why go?
Work side-by-side with top devs and mentors
Use cutting-edge orchestration tools and infra
Compete for up to $10,000 in prizes
Join from anywhere—or find a local hub
Perfect if you’re ready to stop talking and start building.
🌍 Global: online + local hubs
🛠️ July 11–14
📌 OpenAI Bags $200M Pentagon Contract Under New 'Gov' Arm
Image Credit: Carlos Barria | Reuters
OpenAI just announced “OpenAI for Government” and revealed a new $200M contract with the Department of Defense, its first official deal with the Pentagon. The initiative folds existing partnerships with NASA, NIH, Treasury, and Air Force Labs into one centralized operation targeting U.S. government agencies.
Key points:
One-year pilot with the DoD will focus on AI prototypes for “warfighting and enterprise” needs
ChatGPT Enterprise will be deployed to assist with benefits navigation and administrative tasks for military personnel
Custom models will also support cyber defense and classified use cases, within OpenAI’s stated usage policies
The initiative is headquartered in the Washington, D.C. area, signaling deeper federal integration
AI is being operationalized at the highest levels of state power. While the public debates alignment and misuse, the Pentagon is already prototyping frontier models for combat support. What started as an open research lab is now embedded in the infrastructure of military AI.
💥Gemini 2.5: Google Goes GA with Flash-Lite
Image source: Google
Google has officially stabilized the Gemini 2.5 Pro and Flash models after months in preview while introducing Flash-Lite, a new ultra-fast, cost-efficient version aimed at high-throughput tasks like summarization and classification.
What’s new:
2.5 Flash-Lite launches in preview with the lowest latency and cost of the Gemini family, outperforming previous Lite versions on benchmarks.
2.5 Flash is now fully GA, with simplified pricing: $0.30/M input (up), $2.50/M output (down), and no more confusion between “thinking” vs “non-thinking” tiers.
2.5 Pro continues growing rapidly, especially in dev tools like Cursor and Replit, and is now fully stable at its 06-05 version.
All three models now support “thinking budgets” that let developers control how much reasoning the model applies.
With Flash-Lite, Google is clearly optimizing for volume: less intelligence, more scale. Meanwhile, Pro is being positioned as the go-to for deep reasoning and agents. Gemini is now a complete stack, and the roadmap is pointing beyond Pro toward something bigger. The previews are over. The real competition begins now.
✝️ Pope Leo Targets AI as a Threat to Humanity
Image Credit: WSJ
Pope Leo XIV, the first American pontiff, is making AI’s threat to human dignity a signature issue of his papacy. Echoing the labor rights stance of his namesake Leo XIII during the Gilded Age, Leo XIV is positioning the Catholic Church as a counterforce to unchecked technological power.
What’s happening:
In his first major address, the pope cited AI as a challenge to justice, labor, and human dignity, likening it to a new industrial revolution.
Despite years of tech giants courting the Vatican from Google to Microsoft, the Church is doubling down on calls for a binding global AI treaty, which most CEOs strongly oppose.
This week, executives from Meta, IBM, Palantir, Anthropic, and others are in Rome for a high-stakes Vatican summit on AI ethics and governance.
The new pope, a math graduate from Chicago, is seen as more tech-savvy and more skeptical than his predecessor.
Tech firms once treated the Vatican as a soft-power ally, a symbol of moral legitimacy. But under Leo XIV, that bet may backfire. He’s not just offering commentary but also pushing for political intervention, invoking centuries of Church authority to resist what he sees as AI’s threat to the powerless.
🗣️ Google Search Enters The Chat With a Full-on Audio Experience
Image Credit: Google
Google just flipped the switch on Search Live, a new feature in AI Mode that lets users have fluid voice conversations with its generative AI. Users can now tap a “Live” button in the Google app to speak queries aloud and receive AI-generated audio replies, with links displayed for deeper digging.
What’s new:
You can now multitask with voice-driven search while using other apps.
Transcripts are available if you want to switch to text mid-convo.
History is saved, so you can revisit past answers.
The system runs on a custom Gemini model with Google Search’s ranking stack underneath.
Follow-up questions are part of the flow, turning Search into a back-and-forth conversational assistant. Google says the experience will expand soon to handle real-time visual input through your phone camera which is part of its broader push toward multimodal AI search previewed at I/O.
🚀 Boost your business with us—advertise where 10M+ AI leaders engage
🌟 Sign up for the first AI Hub in the world.
