GenAI Daily - February 13, 2026: Perplexity's Model Council, Meta's $10B Data Center, Samsung HBM4 Shipments

Top Story

**

Perplexity launched Model Council, a multi-model research feature that brings several AI models together for one answer. Instead of verifying queries across multiple models manually, Model Council allows users to run the same query across several models at once. When users select Model Council in the main Perplexity interface, their query runs across three of the models available on Perplexity at once, such as Claude Opus 4.6, GPT 5.2, and Gemini 3.0. A synthesizer model reviews the outputs, resolves conflicts where possible, and gives users one answer that shows where the models agree and where they differ.

The feature is available now for Perplexity Max subscribers on the web, with mobile app support coming soon.

Model Council adds multi-model cross-validation as a native UI primitive

, addressing what

the company calls AI model blind spots: "Every AI model has blind spots. It might overlook context, lean toward certain perspectives, or fill gaps with confident guesses. For research you're acting on, it's a big risk."

Why it matters: Model Council represents a shift from single-model reliance to ensemble verification, potentially becoming the standard approach for high-stakes enterprise decision-making where accuracy and confidence matter more than speed.

Key Developments

OpenAI Launches Frontier Enterprise Agent Platform

OpenAI introduced Frontier, a new platform that helps enterprises build, deploy, and manage AI agents that can do real work. Frontier gives agents the same skills people need to succeed at work: shared context, onboarding, hands-on learning with feedback, and clear permissions and boundaries. That's how teams move beyond isolated use cases to AI coworkers that work across the business.

HP, Intuit, Oracle, State Farm, Thermo Fisher, and Uber are among the first to adopt Frontier, and dozens of existing customers - including BBVA, Cisco, and T-Mobile - have already piloted Frontier's approach.

The platform is compatible with agents from third parties like Google, Microsoft and Anthropic.

Impact: Frontier positions OpenAI directly against traditional enterprise software vendors by offering agents that can execute workflows without humans logging into existing SaaS platforms. (OpenAI)

Meta Breaks Ground on $10B Indiana Data Center

Meta is breaking ground on its newest data center: a state-of-the-art 1GW campus in Lebanon, Indiana. This facility represents an investment of over $10 billion in data center infrastructure and the surrounding community, one of Meta's largest infrastructure investments to date.

Designed to deliver 1GW of capacity, this data center will have the flexibility to handle both AI workloads and core products. As AI advances and compute demands continue to grow, gigawatt sites like this one will be critical to advancing the technology that supports core business as well as AI ambitions.

At the peak of construction, this data center is expected to support more than 4,000 construction jobs, and approximately 300 operational jobs. Meta is also launching a new Boone County-wide workforce development program through the Boone County Career Collaborative.

Impact: Meta's 1GW facility signals the infrastructure requirements for AI at scale, representing a fundamental shift from data centers as "backend IT" to strategic assets determining AI capability. (Meta)

Samsung Ships Industry-First Commercial HBM4 Chips

Samsung Electronics announced that it has begun mass production of its industry-leading HBM4 and has shipped commercial products to customers. This achievement marks a first in the industry, securing an early leadership position in the HBM4 market.

Samsung anticipates that its HBM sales will more than triple in 2026 compared to 2025, and is proactively expanding its HBM4 production capacity.

Samsung says HBM4 delivers a consistent 11.7 Gbps transfer rate per pin and can be tuned up to 13 Gbps. The company positions 11.7 Gbps as about 46% higher than an 8 Gbps "industry standard," and also says it is 1.22x higher than the maximum pin speed it cites for HBM3E. Samsung also lists up to 3.3 TB/s of memory bandwidth per stack, which it says is 2.7x higher than HBM3E.

Impact: Samsung's HBM4 leadership could reshape the AI accelerator supply chain, as memory bandwidth increasingly determines training cluster throughput and inference economics. (Samsung)

Funding & Deals

• Apptronik - $520M Series A Extension:

The Austin-based maker of humanoid robots has raised $520 million in a Series A extension round to accelerate production of its human-like robots.

Apptronik secured $520 million to scale humanoid robots for logistics and industrial use.

The company is targeting warehouse and manufacturing applications. (Tech Startups)

• Inertia Enterprises - $450M Series A:

Inertia Enterprises, a fusion energy startup, has secured a massive $450 million Series A to develop one of the world's most powerful laser systems for grid-scale fusion power. Co-founded by Twilio CEO Jeff Lawson alongside leading fusion scientists, Inertia aims to leverage laser inertial confinement technology pioneered at Lawrence Livermore's NIF facility to achieve reliable, net-positive fusion energy. This huge capital injection will enable Inertia to build lasers capable of firing 10-kilojoule pulses at high frequency - a critical step toward its goal of starting construction on a commercial fusion power plant by 2030.

(Tech Startups)

• Resolve AI - $125M Series A:

Resolve AI closed $125M Series A to power autonomous incident detection in enterprise systems

• Building AI-powered systems to detect and resolve enterprise infrastructure issues automatically. (Tech Startups)

• Algorized - $13M Series A:

Swiss-American startup Algorized has secured $13 million in Series A funding to advance its "Physical AI" platform, which gives robots a nervous system for safety and awareness. Algorized is pioneering an edge-native Predictive Safety Engine that uses signals from wireless sensors (like ultra-wideband radio and Wi-Fi) to detect human presence and predict human intent in real time on factory floors. The technology allows industrial robots and machines to operate at full speed around people without the frequent emergency stops required by traditional safety systems.

Led by Run Ventures with Amazon's Industrial Innovation Fund participating. (Tech Startups)

Product Launches

• Perplexity Model Council: Multi-model research feature running three frontier AI models simultaneously (Claude Opus 4.6, GPT-5.2, Gemini 3.0) with synthesized outputs showing agreement and disagreement. Available to Max subscribers on web. (Perplexity)

• Samsung HBM4 Commercial Shipments: Next-generation high-bandwidth memory achieving 11.7-13 Gbps speeds with 2.7x higher bandwidth than HBM3E. First commercial HBM4 shipments globally, targeting AI accelerators from NVIDIA and other partners. (Samsung)

Related reading: Check out this week's [Deep Insights analysis] for strategic context on these developments.