AI Morning Briefing: May 19, 2026
🚀 Google I/O 2026: Gemini 3.5 Flash Launches Live
Google kicked off its annual I/O developer conference today with a major AI release. CEO Sundar Pichai unveiled the Gemini 3.5 family of models, starting with Gemini 3.5 Flash — available immediately at no cost to billions of users worldwide. The new model outperforms Gemini 3.1 Pro on most benchmarks while delivering 4× faster output tokens per second. Google DeepMind CTO Koray Kavukcuoglu noted the model excels at multi-agent deployment and long-running tasks, even successfully building a working operating system from scratch in internal tests. Gemini 3.5 Pro is expected in June. Google also revealed that Gemini now serves 900 million monthly active users, up from 400 million last year. (Mashable | Google Blog)
🥽 Google Previews AI-Powered Smart Glasses with Gemini
In another I/O highlight, Google gave a first look at its audio smart glasses powered by the Gemini assistant. Developed in partnership with Samsung, Warby Parker, and Gentle Monster, the glasses deliver information via audio — spoken privately into the user's ear — rather than through a display. A live demo showed the glasses ordering coffee via DoorDash, reading text messages, and adding calendar events. The glasses will be compatible with both Android and iOS and are slated for a fall 2026 launch. This puts Google in direct competition with Meta, which sold 7 million Ray-Ban smart glasses in 2025. (CNBC)
🤖 Google Introduces Managed Agents in the Gemini API
Google also launched Managed Agents in the Gemini API, enabling developers to spin up an AI agent that reasons, uses tools, and executes code in an isolated Linux environment — all with a single API call. The "Antigravity" agent, powered by Gemini 3.5 Flash, is available immediately via the Interactions API and Google AI Studio. Developers can define custom agents using markdown files (AGENTS.md, SKILL.md), and enterprise support is available through the Gemini Enterprise Agent Platform in private preview. This launch signals Google's push into agent infrastructure, competing with OpenAI's Codex and emerging agent frameworks. (Google Blog)
🔐 OpenAI Launches Daybreak: GPT-5.5 Tiers and Codex Security
OpenAI's Daybreak initiative, announced May 11 and gaining traction this week, introduces a new GPT-5.5 model family with tiered access controls and Codex Security — an agentic harness for application security workflows. The four-stage architecture covers threat modeling, vulnerability discovery, exploitability validation, and automated patch generation. The tiered model system includes GPT-5.5 (standard), GPT-5.5 with Trusted Access for Cyber (verified defensive workflows), and GPT-5.5-Cyber (authorized red teaming). Partners include Akamai, Cisco, Cloudflare, CrowdStrike, NVIDIA, Oracle, and Palo Alto Networks. The launch positions OpenAI to compete for enterprise security budgets and marks a significant expansion beyond coding into the AppSec pipeline. (Futurum Group)
🏥 Cohere Acquires Reliant AI, Expands into Life Sciences
Canadian enterprise AI company Cohere announced today it has acquired Reliant AI, a biopharma analytics platform. The acquisition marks Cohere's entry into the life sciences sector, following its recent merger with Germany's Aleph Alpha to create what the company calls a "transatlantic AI powerhouse." The move is part of a broader consolidation trend in the enterprise AI space, as companies seek to differentiate beyond the general-purpose model market. Financial terms of the acquisition were not disclosed. (Times Colonist | TechCrunch)
📋 Colorado Governor Signs SB 189, Rewriting State AI Law
On May 18, Colorado Governor Jared Polis signed SB 26-189, which repeals and replaces the state's landmark 2024 AI Act with a narrower, more business-friendly framework. The original law's broad "high-risk AI system" and "algorithmic discrimination" framework is replaced with a regime focused on "automated decision-making technology" (ADMT) that processes personal data to materially influence consequential decisions. The new law shifts away from mandatory impact assessments and risk management programs toward targeted consumer disclosures, post-adverse-outcome explanations, correction rights, and meaningful human review. Enforcement is expected to begin once rulemaking is finalized, likely by January 2027. This could serve as a model for other states navigating the tension between AI innovation and consumer protection. (Consumer Finance Monitor | Holland & Knight)
💰 Q1 2026 Shatters AI Funding Records
The AI investment boom continued at unprecedented scale in Q1 2026. According to Crunchbase data cited by the New York Times, AI companies raised a combined $297 billion in the first three months of the year — the strongest quarter since Q2 2022. OpenAI's $122 billion round led the pack, followed by Anthropic's $30 billion Series G, xAI's $20 billion raise, and Waymo's $16 billion. Foundational AI startup funding more than doubled year-over-year. The massive capital inflows are reshaping the venture landscape, with some analysts warning of a "SaaSpocalypse" as traditional SaaS companies face pressure from AI-native alternatives. Tech megacaps are also projected to invest over $300 billion in AI infrastructure this year. (New York Times | Crunchbase)
📌 Quick Hits
- EU AI Act: The August 2026 compliance deadline for high-risk AI systems is fast approaching, with the European Commission's AI Office ramping up enforcement guidance. (AI Act Portal)
- Google Gemini Omni: Google also previewed Gemini Omni, a new multimodal model, at I/O 2026. (Google Blog)
- AI Mode in Search: Google's AI Mode for Search is expanding across the U.S., changing how users query the web. (Google Blog)
For more on AI agent tooling and development workflows, check out NiteAgent — your resource for building and deploying production AI agents.
← Back to all posts
