Meta unveil new AI app at inaugural LlamaCon

Today’s Highlights:

📰 News: Meta unveils new AI app at inaugural LlamaCon

💰 Funding: Mira Murati set to raise $2Bn for her startup

⚡️ Top News Stories:

1. Meta held its inaugural LlamaCon event, where they unveiled both a consumer chatbot app and a streamlined Llama API platform.

• The featured launch is a stand-alone AI assistant app that builds on its existing platform integrations, offering ChatGPT-style functionality enhanced by the wealth of user data it already collects across Facebook and Instagram.

• The app includes a Discover feed allowing users to share AI-generated content with friends, tapping into viral trends while giving users control over what they post publicly. 

• They also showcased developer-friendly preview of the Llama API with open-source flexibility, tooling for fine-tuning and evaluation, and seamless integration with major SDKs to make building and deploying Llama-based applications easier and more secure.

• To protect and scale open-source AI, they also introduced a suite of new safety tools — including Llama Guard 4 and Prompt Guard 2 — and launched the Llama Defenders Program.

• At LlamaCon, Meta launched a new app that lets users interact with Meta AI directly.

2. The Trump administration is considering replacing Biden-era AI chip export tiers with country-specific licensing deals, aiming to enhance U.S. leverage in trade talks while maintaining tighter control over access to powerful semiconductors.

3. Anthropic has launched Integrations and Advanced Research for Claude, enabling users to connect apps like Atlassian and Zapier and conduct deep, citation-backed research across internal and external data sources, as it ramps up efforts to compete with ChatGPT and Gemini.

4. Google has expanded access to its AI Mode search tool for all U.S. adults in Labs and introduced major new features—including visual cards, real-time local info, and search continuity.

5. Amazon has unveiled Nova Premier, its most capable AI model yet, designed for complex multimodal tasks and available on Bedrock, positioning it as a distillation tool for smaller models.

6. Microsoft has launched three new open-source AI models—Phi 4 mini reasoning, Phi 4 reasoning, and Phi 4 reasoning plus—designed specifically for advanced reasoning tasks like math, science, and coding.

7. Sam Altman’s Tools for Humanity has launched Orb Mini, a portable eyeball-scanning device that issues blockchain-based proof-of-human IDs, aiming to expand its World verification network amid concerns over distinguishing humans from AI agents online. They also announced partnerships with Tinder, Visa, and Stripe to expand the use of its iris-based verification system, and unveiled the World Card for crypto spending and enabling logins and payments across platforms like Kalshi and Morpho.

8. Freepik has launched F Lite, a 10-billion-parameter open AI image generator trained solely on licensed, safe-for-work images, offering a legally sound alternative to models trained on copyrighted content amid rising copyright scrutiny.

9. Visa and Mastercard have launched AI shopping agents—Intelligent Commerce and Agent Pay respectively—that enable consumers to delegate purchases to AI based on personal preferences, signaling a broader industry trend toward agentic commerce as seen with similar efforts from PayPal, Amazon, and OpenAI.

10. Google has expanded NotebookLM’s AI-generated Audio Overviews to 76 additional languages and introduced an Output Language setting, allowing users to create personalized, multilingual podcast-style summaries of their uploaded documents for learning and accessibility.

11. A new study by researchers from Cohere, Stanford, MIT, and AI2 accuses LM Arena—the group behind Chatbot Arena—of providing top AI firms like Meta, OpenAI, Google, and Amazon with privileged private testing access, allowing them to improve leaderboard scores while excluding smaller competitors, raising ethical concerns about benchmark integrity as LM Arena prepares to commercialize its operations

12. Ai2 has released Olmo 2 1B, a fully open-source, 1-billion-parameter AI model that outperforms Google, Meta, and Alibaba's small models on reasoning and factuality benchmarks, while remaining lightweight enough to run on consumer hardware.

13. DeepSeek has released Prover V2, an advanced math-focused AI built on its 671B-parameter MoE-based V3 model, as part of a broader push to strengthen its lineup of specialized and reasoning-capable tools in the open-source AI space.

14. FutureHouse, a nonprofit backed by Eric Schmidt, has released its first suite of AI-powered tools—Crow, Falcon, Owl, and Phoenix—aimed at supporting scientific research through literature search, experiment planning, and subject-area mapping.

• Each AI tool is designed for a specific task: Crow answers questions about scientific papers, Falcon performs deep literature/database searches, Owl finds related work in scientific domains, and Phoenix helps design chemistry experiments.

15. Duolingo has launched 148 new beginner-level courses powered by generative AI, enabling users of all 28 supported languages to learn top global languages like Japanese, Korean, and Mandarin, in the company’s most expansive update to date.

16. North Korea has infiltrated hundreds of Western companies by placing fake IT workers under stolen identities—often aided by U.S.-based facilitators and AI tools—in a sophisticated scheme that’s quietly funneled millions into Pyongyang’s coffers while evading traditional corporate vetting systems.

17. At a Washington tech conference, Nvidia CEO Jensen Huang warned that China is close behind the U.S. in AI, praised Huawei as one of the world’s most formidable tech companies, and urged U.S. policymakers to focus on domestic innovation rather than relying on restrictive chip export controls.

💰 Top Funding News:

1. Mira Murati, former CTO of OpenAI and now CEO of Thinking Machines Lab, is orchestrating a significant funding round led by Andreessen Horowitz, aiming to raise $2Bn at a $10Bn valuation, while retaining an unusually high level of control over the company.

2. Astronomer, which powers AI-driven data orchestration and observability through its Astro platform, raised a $93M Series D led by Bain Capital Ventures, with participation from Salesforce Ventures, Insight, Meritech, Venrock, and Bosch Ventures.

3. Cast AI, which automates Kubernetes cloud operations with AI to optimize application performance and infrastructure costs, raised a $108M Series C led by G2 Venture Partners and SoftBank Vision Fund 2, with participation from Aglaé Ventures, Hedosophia, Cota Capital, Vintage Investment Partners, Creandum, and Uncorrelated Ventures.

4. Supio, which combines legal AI and expert verification to streamline litigation and enhance outcomes in personal injury and mass tort cases, raised a $60M Series B led by Sapphire Ventures, with participation from Mayfield and Thomson Reuters Ventures.

5. Gruve.ai, which provides an outcome-based AI services platform to operationalize AI across enterprise environments, raised a $20M Series A led by Mayfield, with participation from Cisco Investments and other investors, bringing total funding to $37.5M.

6. Dex, which uses generative AI voice and machine learning to match candidates with aligned job opportunities and streamline hiring, raised a $3.1M Pre-Seed round led by a16z speedrun and Concept Ventures.

That's all for today's email! If you want more please follow us at the social channels linked below, or check out our website!

Share our newsletter: If you like our work please share/forward this email with your friends, colleagues, and family. It's the best way to support us!

If this email was forwarded to you please sign up here to continue receiving them.

Want your content, product, jobs, or event featured in our newsletter? Reply to this email with the details, and our team will reach out to you.

Do you use AI for work? Tell us how, and you could be featured in our newsletter!

Check out our website for more resources, including a list of AI investors, products, events, and twitter follows.

For an archive of all our posts, click here.

We'd love to hear from you! You can always leave us comments or feedback by replying to this email!

Powered by AI. Curated and edited by Humans.