For the past year, the enterprise AI community has been locked in a debate about how much freedom to give AI agents. Too little, and you get expensive workflow automation that barely justifies the “agent” label. Too much, and you get the kind of data-wiping disasters that plagued early adopters of tools like OpenClaw. This week, Google Labs released an update to Opal, its no-code visual agent builder, that quietly lands on an answer — and it carries lessons that every IT leader planning an agent strategy should study carefully.

The update introduces what Google calls an “agent step” that transforms Opal’s previously static, drag-and-drop workflows into dynamic, interactive experiences. Instead of manually specifying which model or tool to call and in what order, builders can now define a goal and let the agent determine the best path to reach it — selecting tools, triggering models like Gemini 3 Flash or Veo for video generation, and even initiating conversations with users when it needs more information.

It sounds like a modest product update. It is not. What Google has shipped is a working reference architecture for the three capabilities that will define enterprise agents in 2026:

1. Adaptive routing

2. Persistent memory

3. Human-in-the-loop orchestration

…and it’s all made possible by the rapidly improving reasoning abilities of frontier models like the Gemini 3 series.

The ‘off the rails’ inflection point: Why better models change everything about agent design

To understand why the Opal update matters, you need to understand a shift that has been building across the agent ecosystem for months.

The first wave of enterprise agent frameworks — tools like the early versions of CrewAI and the initial releases of LangGraph — were defined by a tension between autonomy and control. Early models simply were not reliable enough to be trusted with open-ended decision-making. The result was what practitioners began calling “agents on rails”: tightly constrained workflows where every decision point, every tool call, and every branching path had to be pre-defined by a human developer.

This approach worked, but it was limited. Building an agent on rails meant anticipating every possible state the system might encounter — a combinatorial nightmare for anything beyond simple, linear tasks. Worse, it meant that agents could not adapt to novel situations, the very capability that makes agentic AI valuable in the first place.

The Gemini 3 series, along with recent releases like Anthropic’s Claude Opus 4.6  and Sonnet 4.6, represents a threshold where models have become reliable enough at planning, reasoning, and self-correction that the rails can start coming off. Google’s own Opal update is an acknowledgment of this shift. The new agent step does not require builders to pre-define every path through a workflow. Instead, it trusts the underlying model to evaluate the user’s goal, assess available tools, and determine the optimal sequence of actions dynamically.

This is the same pattern that made Claude Code’s agentic workflows and tool calling viable: the models are good enough to decide the agent’s next step and often even to self-correct without a human manually re-prompting every error. The difference compared to Claude Code is that Google is now packaging this capability into a consumer-grade, no-code product — a strong signal that the underlying technology has matured past the experimental phase.

For enterprise teams, the implication is direct: if you are still designing agent architectures that require pre-defined paths for every contingency, you are likely over-engineering. The new generation of models supports a design pattern where you define goals and constraints, provide tools, and let the model handle routing — a shift from programming agents to managing them.

Memory across sessions: The feature that separates demos from production agents

The second major addition in the Opal update is persistent memory. Google now allows Opals to remember information across sessions — user preferences, prior interactions, accumulated context — making agents that improve with use rather than starting from zero each time.

Google has not disclosed the technical implementation behind Opal’s memory system. But the pattern itself is well-established in the agent-building community. Tools like OpenClaw handle memory primarily through markdown and JSON files, a simple approach that works well for single-user systems. Enterprise deployments face a harder problem: maintaining memory across multiple users, sessions, and security boundaries without leaking sensitive context between them.

This single-user versus multi-user memory divide is one of the most under-discussed challenges in enterprise agent deployment. A personal coding assistant that remembers your project structure is fundamentally different from a customer-facing agent that must maintain separate memory states for thousands of concurrent users while complying with data retention policies.

What the Opal update signals is that Google considers memory a core feature of agent architecture, not an optional add-on. For IT decision-makers evaluating agent platforms, this should inform procurement criteria. An agent framework without a clear memory strategy is a framework that will produce impressive demos but struggle in production, where the value of an agent compounds over repeated interactions with the same users and datasets.

Human-in-the-loop is not a fallback — it is a design pattern

The third pillar of the Opal update is what Google calls “interactive chat” — the ability for an agent to pause execution, ask the user a follow-up question, gather missing information, or present choices before proceeding. In agent architecture terminology, this is human-in-the-loop orchestration, and its inclusion in a consumer product is telling.

The most effective agents in production today are not fully autonomous. They are systems that know when they have reached the limits of their confidence and can gracefully hand control back to a human. This is the pattern that separates reliable enterprise agents from the kind of runaway autonomous systems that have generated cautionary tales across the industry.

In frameworks like LangGraph, human-in-the-loop has traditionally been implemented as an explicit node in the graph — a hard-coded checkpoint where execution pauses for human review. Opal’s approach is more fluid: the agent itself decides when it needs human input based on the quality and completeness of the information it has. This is a more natural interaction pattern and one that scales better, because it does not require the builder to predict in advance exactly where human intervention will be needed.

For enterprise architects, the lesson is that human-in-the-loop should not just be treated as a safety net bolted on after the agent is built. It should be a first-class capability of the agent framework itself — one that the model can invoke dynamically based on its own assessment of uncertainty.

Dynamic routing: Letting the model decide the path

The final significant feature is dynamic routing, where builders can define multiple paths through a workflow and let the agent select the appropriate one based on custom criteria. Google’s example is an executive briefing agent that takes different paths depending on whether the user is meeting with a new or existing client — searching the web for background information in one case, reviewing internal meeting notes in the other.

This is conceptually similar to the conditional branching that LangGraph and similar frameworks have supported for some time. But Opal’s implementation lowers the barrier dramatically by allowing builders to describe routing criteria in natural language rather than code. The model interprets the criteria and makes the routing decision, rather than requiring a developer to write explicit conditional logic.

The enterprise implication is significant. Dynamic routing powered by natural language criteria means that business analysts and domain experts — not just developers — can define complex agent behaviors. This shifts agent development from a purely engineering discipline to one where domain knowledge becomes the primary bottleneck, a change that could dramatically accelerate adoption across non-technical business units.

What Google is really building: An agent intelligence layer

Stepping back from individual features, the broader pattern in the Opal update is that Google is building an intelligence layer that sits between the user’s intent and the execution of complex, multi-step tasks. Building on lessons from an internal agent SDK called “Breadboard”, the agent step is not just another node in a workflow — it is an orchestration layer that can recruit models, invoke tools, manage memory, route dynamically, and interact with humans, all driven by the ever improving reasoning capabilities of the underlying Gemini models.

This is the same architectural pattern emerging across the industry. Anthropic’s Claude Code, with its ability to autonomously manage coding tasks overnight, relies on similar principles: a capable model, access to tools, persistent context, and feedback loops that allow self-correction. The Ralph Wiggum plugin formalized the insight that models can be pressed through their own failures to arrive at correct solutions — a brute-force version of the self-correction that Opal now packages some of that into a polished consumer experience.

For enterprise teams, the takeaway is that agent architecture is converging on a common set of primitives: goal-directed planning, tool use, persistent memory, dynamic routing, and human-in-the-loop orchestration. The differentiator will not be which primitives you implement, but how well you integrate them — and how effectively you leverage the improving capabilities of frontier models to reduce the amount of manual configuration required.

The practical playbook for enterprise agent builders

Google shipping these capabilities in a free, consumer-facing product sends a clear message: the foundational patterns for building effective AI agents are no longer cutting-edge research. They are productized. Enterprise teams that have been waiting for the technology to mature now have a reference implementation they can study, test, and learn from — at zero cost.

The practical steps are straightforward. First, evaluate whether your current agent architectures are over-constrained. If every decision point requires hard-coded logic, you are likely not leveraging the planning capabilities of current frontier models. Second, prioritize memory as a core architectural component, not an afterthought. Third, design human-in-the-loop as a dynamic capability the agent can invoke, rather than a fixed checkpoint in a workflow. And fourth, explore natural language routing as a way to bring domain experts into the agent design process.

Opal itself probably won’t become the platform enterprises adopt. But the design patterns it embodies — adaptive, memory-rich, human-aware agents powered by frontier models — are the patterns that will define the next generation of enterprise AI. Google has shown its hand. The question for IT leaders is whether they are paying attention.

Alongside the investment, OpenAI and Amazon have also reached a deal in which OpenAI will utilise 2GW of computing capacity powered by Amazon’s in-house Trainium chips.

US artificial intelligence platform OpenAI has today (27 February) announced a $110bn funding round, reportedly a record for the private technology company, with a figure that is double that of its previous funding round. 

Amazon invested $50bn, Nvidia invested $30bn and SoftBank invested $30bn. The investment brings OpenAI from a $500bn valuation to a $730bn pre-money valuation and OpenAI has stated that the organisation expects additional investors to join as the round progresses. 

Commenting on the news, Sam Altman, the CEO of OpenAI, told CNBC: “We’re super excited about this deal. AI is going to happen everywhere. It’s transforming the whole economy and the world needs a lot of collective computing power to meet the demand.”

OpenAI also confirmed that the funding announcement will not impact the terms of a current partnership it holds with tech giant Microsoft, which was established in 2019. In a joint statement, both organisations agreed that the deal is “strong and central” to operations. CNBC also reported that Microsoft has the option to participate in OpenAI’s funding round. 

Amazon’s $50bn investment in OpenAI will begin with an initial commitment of $15bn, followed by another $35bn over the course of the next few months when “certain conditions” are met. OpenAI also has an additional deal with Amazon in which the organisation will utilise 2GW of computing capacity powered by Amazon’s in-house Trainium chips.

The announcement comes at a time when many of the globe’s most influential technology companies are vying for dominance in the AI space. Earlier this week, Intrinsic, an Alphabet-owned software and AI company, announced that it was joining Google, as a means of enabling Google to move further into the physical AI space.  

Japan’s SoftBank was also recently announced as the first to deploy SambaNova’s new SN50 chips within its data centres in Japan, while Intel is also partnering with SambaNova to roll out an Intel-powered AI cloud.

Don’t miss out on the knowledge you need to succeed. Sign up for the Daily Brief, Silicon Republic’s digest of need-to-know sci-tech news.

Sick of your smart ring’s battery not holding up? Ultrahuman’s new $479 Ring Pro smart ring, unveiled on Friday, offers up to 15 days of battery life on a single charge. The Ring Pro joins the company’s $349 Ring Air, which boosts health tracking, thanks to longer battery life, increased data storage, improved speed and accuracy and a new heart-rate sensing architecture. The ring works in conjunction with the latest Pro charging case. 

Ultrahuman also launched its Jade AI, which can act as an agent based on analysis of current and historical health data. Jade can synthesize data from across the company’s products and is compatible with its Rings.

“With industry-leading hardware paired with Jade biointelligence AI, users can now take real-time actionable interventions towards their health than ever before,” said Mohit Kumar, CEO of Ultrahuman.

No US sales

That hardware isn’t available in the US, though, thanks to the ongoing ban on Ultrahuman’s Rings sales here, stemming from a patent dispute with its competitor, Oura Ring. It’s available for preorder now everywhere else and is slated to ship in March. Jade’s available globally.

Ultrahuman says the Ring Pro boosts battery life to about 15 days in Chill mode — up to 12 days in Turbo — compared to a maximum of six days for the Air. The Pro charger’s battery stores enough for another 45 days, which you top off with Qi-compatible wireless charging. In addition, the case incorporates locator technology via the app and a speaker, as well as usability features such as haptic notifications and a power LED.

The ring can also retain up to 250 days of data versus less than a week for the cheaper model. Ultrahuman redesigned the heart-rate sensor for better signal quality. An upgraded processor improves the accuracy of the local machine learning and overall speed. 

It’s offered in gold, silver, black and titanium finishes, with available sizes ranging from 5 to 14.

Jade’s Deep Research Mode is the cross-ecosystem analysis feature, which aggregates data from Ring and Blood Vision and the company’s subscription services, Home and M1 CGM, to provide historical trends, offer current recommendations and flag potential issues, as well as trigger activities such as A-fib detection. Ultrahuman plans to expand its capabilities to include health-adjacent activities, such as ordering food.

Some new apps are also available for the company’s PowerPlug add-on platform, including capabilities such as tracking GLP-1 effects, snoring and respiratory analysis and migraine management tools.

Apple’s rumored Studio Display refresh is back in the spotlight. While earlier reports suggested the company had two new models in the pipeline, fresh details (via Macworld) now hint at what could actually change. Newly uncovered code and leaks point toward upgrades to ports and speakers, offering the clearest picture yet of how Apple might evolve its pro-focused monitor lineup.

Just to refresh your memory, Apple introduced the original Studio Display in 2022 as a 27-inch 5K monitor designed to pair with Macs, featuring a built-in camera, speakers, and Thunderbolt connectivity. While the display has remained a popular choice for Mac users, it has seen few hardware changes since launch. That makes the signs of a refresh particularly noteworthy.

A refresh focused on ports and sound

First, the new models are expected to bring upgraded ports. As Macs continue adopting faster Thunderbolt and USB standards, improved connectivity would help the Studio Display better match modern workflows. Faster ports could support higher bandwidth for accessories, external storage, and multi-display setups, making the monitor more capable as a desktop hub.

The second rumored upgrade centers on improved speakers. The current Studio Display already includes a six-speaker sound system, but Apple is reportedly testing enhanced audio hardware for the new models. Better speakers could make the display even more suitable for video editing, music playback, and video calls without requiring external speakers.

As for how the two models might differ, nothing has been confirmed yet. Still, it would be unusual for Apple to release two nearly identical 27-inch displays with only minor changes to ports or speakers, which has led to speculation that one could be a larger 32-inch variant. For now, Apple has not officially acknowledged these displays, so it is best to treat the rumors with caution. That said, if the reports do prove to be accurate, Mac users who have been waiting for a Studio Display refresh may not have to wait much longer.