Walmart rolls out AI-powered super agents, OpenAI launches ChatGPT Agent: this week's news roundup

From the US AI Action Plan to the launch of more agentic AI from OpenAI and Walmart, here’s a roundup of the news that caught our attention this week.

Trump launches US AI Action Plan

President Trump’s AI Action Plan, the replacement to Biden’s 2023 AI Executive Order, was officially launched this week. The plan lays out three central pillars for furthering US leadership in AI: accelerating innovation, building out American infrastructure and leading international diplomacy and security. It’s essentially a shift from the guardrail-heavy policy under Biden to rapid infrastructure build‑out, deregulation, and global expansion of US AI leadership. 

Critics are already voicing concerns about the action plan. According to an article in Wired: “A variety of tech, consumer protection, and civil society organisations are already speaking out with concerns about the action plan. Sarah Myers West and Amba Kak, the co-executive directors of the AI Now Institute, characterised the plan in a statement to Wired as ‘written by Big Tech interests invested in advancing AI that's used on us, not by us.’”

Walmart rolls out AI-powered ‘super agents’

US retailer Walmart is to roll out a suite of AI-powered super agents in a push towards personalised shopping, internal service, seller onboarding, and developer tools being delivered by agentic AI. 

There are four super agents – one for customers (Sparky), one for employees (Associate), one for engineers (Developer), and one for sellers and suppliers (Marty). These super agents consolidate a number of purpose-built agents under a unified interface tailored to its user group. Previously, users navigated a number of siloed tools. Sparky is already live and the other three agents will be launched over the next year.

Walmart told the Wall Street Journal the super agents will simplify the user experience, with John Furner, CEO of Walmart’s US Division, saying he “believes AI agents will help deliver top-line growth, as they give customers more personalized and enticing shopping experiences, as well as bottom-line savings, where they can help manage supply chains and inventory more efficiently, among other areas”.

Amazon buys wearables startup Bee

Amazon has bought Bee, a US startup that has developed a Fitbit-like wearable that passively listens to and records everything you say. Terms of the deal have not been disclosed, and all Bee staff are thought to be joining Amazon.

The Bee wearable device listens to your conversations and then uses AI to transcribe everything said by you and the people around you. It can then generate summaries, reminders and suggestions.

It’s not the first time Amazon has entered the wearables market. It still sells Echo Frames smart glasses, but its Halo line of health and fitness trackers was discontinued in 2023 after poor takeup. Significantly, Bee is not a fitness tracker – rather, it adds to Amazon's ambitions in personal, contextual AI, and complements its investments in Alexa and AWS AI tools. 

OpenAI releases ChatGPT Agent and agrees UK government deal

OpenAI has launched ChatGPT Agent, combining earlier tools (Operator and Deep Research) into one autonomous assistant that can "think and act" on your behalf. While still in beta, OpenAI says ChatGPT Agent means “ChatGPT can now do work for you using its own computer, handling complex tasks from start to finish”, adding, “ChatGPT will intelligently navigate websites, filter results, prompt you to log in securely when needed, run code, conduct analysis, and even deliver editable slideshows and spreadsheets that summarize its findings”.

As Aaron Levie, CEO at Box, commented on LinkedIn: “ChatGPT just launched an agent that has full access to its own computer. This means that it can access far more tools and run programs in the background to do work. When combined with your existing corporate information, this becomes insanely powerful for doing knowledge work.”

OpenAI has also agreed to a partnership with the UK government. The memorandum of understanding (MoU) signed this week is aimed at boosting public sector productivity and builds on the UK’s AI Opportunities Action Plan. The UK government will explore integrating OpenAI's GPT models into public services like justice, education, civil service, tax, and defence. OpenAI will scale its London operations to support research, engineering, and public deployment and explore investment in UK-based AI infrastructure. OpenAI will also collaborate with the UK’s AI Security Institute to expand government knowledge of model capabilities and support AI safety research.

The deal has already been criticised for its vagueness and lack of transparency. The MoU is non-binding and contains no legal enforcement mechanism, with a CityAM report saying that civil liberties and digital rights groups warn that the MoU may move “too fast without democratic input”.

YouTube launches creation tools for Shorts

YouTube has unveiled a series of AI-powered creation tools for short-form content (YouTube Shorts), designed to make video production faster and accessible to a broader range of creators. They include photo to video generation, a library of AI-powered visual effects and an AI playground hub that puts creation tools like video, music and images all in one place. It’s free and rolls out over the next week in the US, Canada, Australia, and New Zealand, with more regions added later this year.

Lovable becomes fastest growing software company ever

Mere days after reaching a landmark $2 billion valuation, eight-month old Swedish AI startup Lovable, whose “vibe coding” tech enables users to create apps and websites through simple text prompts, has attained another landmark – $100 million in annual recurring revenue (ARR). This makes it the fastest growing software company ever.

TechCrunch reports that Lovable claims to now have more than 2.3 million active users, and just 45 full-time employees, with 14 open positions on its careers page.