Agentic AI Weekly | Berkeley RDI | February 18, 2026
AgentX–AgentBeats Phase 1 Winners, Phase 2 Preview, Berkeley Xcelerator, Agentic AI Summit, new Lambda Custom Track Workshop
AgentX–AgentBeats Highlights: Phase 1 Winners Announced, Phase 2 Preview Released
Thank you to everyone who participated in Phase 1 of the AgentX–AgentBeats Competition! The judges were amazed by the impressive number of submissions and the exceptional creativity, rigor, and thoughtfulness showcased across all projects.
After an incredibly competitive round of judging, we’re excited to reveal the winners of Phase 1!
Thank you once again for all of your hard work in Phase 1, and we can’t wait to see what you bring forward in Phase 2!
Phase 2 Purple Agent Competition Preview
We're excited to announce that Phase 2 of the AgentX–AgentBeats competition will officially launch on February 23, 2026. Participants will build purple agents to tackle the select top green agents from Phase 1 and compete on the public leaderboards.
Unlike Phase 1, where participants competed across all tracks throughout the entire duration, Phase 2 introduces a sprint-based format. The competition will be organized into three rotating sprints:
First Sprint: Research Agent, Web & Computer-Use Agent, and Game Agent tracks
Second Sprint: Agent Safety, Cybersecurity & Software Testing Agent, and Finance Agent tracks
Third Sprint: Multi-Agent Evaluation, Coding Agent, and Healthcare Agent tracks
For the specific sprint start and end dates, please refer to the AgentX–AgentBeats website for full details. We’ll release the official benchmarks and selected green agents shortly before each sprint begins.
Keep an eye on our announcements, as we may introduce additional tracks throughout the competition based on community interest and emerging opportunities.
Over the next week, Berkeley RDI is holding two new workshops for AgentX-AgentBeats—packed with practical guidance to elevate your AgentX‑AgentBeats projects (Phase 2) and meaningful insights for the broader agentic AI community.
Tomorrow, February 19, Meta, HuggingFace, and Unsloth will be hosting a workshop for the OpenEnv Challenge: a new initiative focused on building open-source reinforcement learning (RL) environments that help drive general intelligence. The workshop, in partnership with the Linux/PyTorch Foundation, will give you expert guidance and an overview of the OpenEnv Challenge, and we hope to see you there!
On February 24, Lambda is hosting a brand new onboarding and strategies session for those interested in Phase 2 of the Lambda Agent Security custom track!
Here’s what you can expect from the session:
Lambda Custom Track Phase 2 Overview: A breakdown of the attackers vs. defenders format, leaderboard mechanics, scoring criteria, and timeline.
Strategic Guidance: How to approach offensive vs. defensive agent design, common failure modes, and how to optimize for performance under realistic constraints.
Security Framework Deep Dive: How the adversarial evaluation framework works and what makes a strong, rigorous submission.
Platform Walkthrough: Registration steps, API access, environment setup, and workflows specific to the Lambda track.
Live Q&A: Clarify rules, strategy, and evaluation details directly with the organizers.
You can click the button below to book your spot, and click here to get an overview of the Lambda Custom Track!
As of today, 1300+ teams from around the world have joined AgentX-AgentBeats. Hosted by Berkeley RDI, this global challenge builds on our amazing Agentic AI MOOC community of ~40K learners, uniting builders, researchers, engineers, and AI enthusiasts worldwide to build, benchmark, and push the boundaries of agentic AI.
Agentic AI Summit 2026 (Early-Bird Pricing and CFP are Live!)
Save the date! The Agentic AI Summit returns to Berkeley on August 1–2, 2026, welcoming 5,000+ expected in-person attendees for two days of insights and innovation. Building on last year’s sold-out success—with 2,000+ in‑person attendees and 40,000+ global livestream participants—the summit will bring together researchers, builders, industry leaders, and the global agentic AI community for keynotes, technical talks and panels, hands-on workshops, live demos, and more!
🎟️ Early‑Bird Pricing (Limited Capacity)
A limited number of early‑bird tickets are still available:
Student Early-Bird: $99
Standard Early-Bird: $249
If you’re looking to secure the best ticket price and be part of the conversation shaping the future of Agentic AI, we encourage you to register early. We look forward to welcoming you to Berkeley this August.
We’re also thrilled to share that the Call for Speaking Proposals (CFP) for the Agentic AI Summit 2026 is now open!
If you’re interested in sharing your work through a technical talk, panel discussion, workshop, or tutorial, or poster presentation—and helping advance the frontiers of the Agentic AI—we warmly invite you and/or your team to apply and be part of the conversation at the Summit.
Please complete the form below to submit your proposal. The program committee will review submissions on a rolling basis.
Berkeley Xcelerator — Applications Now Open!
The Berkeley Xcelerator, a non-dilutive accelerator program designed to support pre-seed and seed-stage startups building at the forefront of Agentic AI, is now open for applications!
The Xcelerator is built in partnership with Berkeley RDI’s research community and ecosystem partners, offering selected teams the support, resources, and guidance to take their startup to the next level! In addition, the Xcelerator is open to everyone; you do not need to be affiliated with UC Berkeley to apply!
Why apply to the Xcelerator?
Unparalleled access to frontier research and expertise through close collaboration with Berkeley RDI’s community across agentic AI, AI safety and security, and the broader AI landscape.
Practical enablement through industry partnerships, including cloud, GPU, and API credits provided by industry partners such as Google Cloud, Google DeepMind, OpenAI, and Nebius, with more to be announced!
Visibility and network effects through the Berkeley ecosystem and Berkeley RDI’s global community of 56,000+ developers and builders, including the rapidly growing Agentic AI MOOC community
A culminating Demo Day at the Agentic AI Summit (August 1–2, 2026), bringing together 5,000+ in-person attendees and placing your startup directly in front of top-tier VCs, leading AI researchers, industry executives, and strategic partners.
We’re looking for AI and Agentic AI startups at the pre-seed or seed stage. If you think that you or your team are a good fit, we encourage you to learn more and apply via the Xcelerator website and form below!
📅 Applications close at the end of February!
Our sincerest thanks to all of our sponsors and partners:
Trends This Week
Anthropic raised $30 billion in a Series G funding round that values the company at $380 billion post-money, with participation from investors including GIC, Coatue, D. E. Shaw Ventures, and Founders Fund. Anthropic stated that its run-rate revenue has reached $14 billion, growing more than 10x annually over each of the past three years, while the number of customers spending over $100,000 annually on Claude has increased sevenfold over the past year. The company said that the new capital, which also includes previously announced investments from Microsoft and Nvidia, will be used to expand infrastructure and make Claude more broadly available to enterprise customers.
OpenAI released a research preview of GPT-5.3-Codex-Spark, a smaller version of GPT-5.3-Codex and its first model specifically designed for real-time coding. Codex-Spark is optimized for ultra-low latency interaction, delivering more than 1,000 tokens per second while enabling developers to make targeted edits and iterate quickly inside Codex. The model runs on Cerebras’ Wafer Scale Engine 3, marking the first OpenAI model to be deployed on a custom Cerebras chip, and is intended to explore new interaction patterns for coding workflows, with initial access available to ChatGPT Pro users while infrastructure scales.
The AI infrastructure boom is affecting hardware supply chains as demand for memory-intensive systems increases pressure on global DRAM production. In a recent Bloomberg report, it was noted that hyperscalers, including Google, OpenAI, and other data-center operators, are purchasing large volumes of Nvidia accelerators, contributing to tighter memory supply and a roughly 75% increase in one category of DRAM prices within a single month. The report also cited comments from Tim Cook, who said the shortage could compress iPhone margins, and Elon Musk, who said Tesla may need to build its own memory fabrication plant in response to supply constraints.
Google DeepMind announced advances in mathematics research through an updated version of Gemini 3 Deep Think, which adds parallel reasoning so the model can explore multiple hypotheses before selecting a solution, improving performance on difficult reasoning benchmarks such as ARC-AGI-2. The company also introduced Aletheia, a math research agent powered by Deep Think that generates and verifies long-horizon proofs. Google DeepMind said the system has already contributed to research-level outcomes and solving open Erdős Problems.
OpenClaw, formerly known as Clawdbot and Moltbot, is being transitioned into an independent foundation, allowing the project to remain open-source while receiving ongoing support from OpenAI. Announcing the change, OpenAI CEO Sam Altman said the future of AI will be “extremely multi-agent,” framing open-source collaboration as an important part of that direction. Altman said personal agents are expected to become a core part of OpenAI’s product offerings, and OpenClaw creator Peter Steinberger will join the company and work on its future agentic systems.
Don’t miss the developments shaping Agentic AI. Subscribe for weekly coverage of groundbreaking research, emerging trends, and critical insights across Agentic AI and the broader AI landscape.
