According to a report by Bloomberg, the upgrade will arrive with iOS 27, iPadOS 27 and macOS 27, expected later this year.

People familiar with the plans said the feature will allow Siri to process multi-step commands in a single query, instead of handling them one at a time.

Right now, Siri responds to one instruction per request. That has left it trailing competitors that can manage more complex tasks in a single interaction. With this change, a user could ask Siri to get directions and share them with a contact in one sentence.

Apple is also working on a comprehensive redesign of Siri. The company is said to be building a more advanced version of the assistant using technology linked to Alphabet Inc.’s Gemini model. Apple has not responded to requests for comment.

The upgrade is expected to feature at the Worldwide Developers Conference on June 8, 2026, where Apple usually previews its next software updates.

Beyond handling multiple requests, Apple is testing a new Siri app with both voice and text input. Users may also be able to revisit past conversations, a feature already common with tools like ChatGPT.

There are also plans for an “Extensions” system that would allow third-party services to plug directly into Siri.

At the same time, Apple is looking at opening Siri to other AI providers. Reports say users could choose between different assistants, including those from Anthropic, alongside existing integrations.

This changes the tech giant’s approach. Apple’s earlier Apple Intelligence rollout in 2024 did not gain strong traction, and the company has been under pressure to close the gap with competing systems.

Internally, the project to overhaul Siri into a full chatbot is said to carry the codename “Campos”. The plan is to embed it across the iPhone, iPad and Mac, replacing the current interface with something more interactive and capable.

The plan, which will begin construction in 2026, aims to add nearly 1.3 gigawatts of computing power, designed to enhance missions ranging from cybersecurity to drug discovery.

“This investment removes the technology barriers that have held government back,” AWS CEO Matt Garman said. “We’re giving agencies expanded access to advanced AI capabilities that will enable them to accelerate critical missions from cybersecurity to drug discovery. This investment removes the technology barriers that have held government back and further positions America to lead in the AI era.”

The new infrastructure will span AWS Top Secret, AWS Secret, and GovCloud regions, providing federal agencies with access to a suite of AI services including Amazon SageMaker for model training and customisation, Amazon Bedrock for deploying models and agents, Amazon Nova, and Anthropic’s Claude chatbot. 

The rollout will also integrate NVIDIA AI chips and AWS Trainium hardware, allowing agencies to run complex simulations, analyse massive datasets, and develop tailored AI solutions with outstanding speed.

AWS has been a long-standing partner of the U.S. government. Its first government cloud infrastructure dates back to 2011, followed by AWS Top Secret-East in 2014, the first air-gapped commercial cloud for classified workloads. 

AWS Secret Region was launched in 2017, accredited to handle all levels of classified data. Today, the company serves over 11,000 federal agencies, making it the largest provider of secure cloud services to the government.

While Amazon tops competitors in the cloud market, companies like Google, Oracle, and Microsoft have been expanding their AI products and services to federal customers. 

Google’s “Google for Government” and enterprise tiers from OpenAI and Anthropic have all been offered at nominal rates to encourage adoption.

Experts also noted the geopolitical implications. “The U.S. is in an AI arms race with China and will significantly increase its AI compute capacity to maintain its lead,” said D.A. Davidson analyst Gil Luria. 

AWS’s new infrastructure will allow agencies to convert previously fragmented intelligence, supply chain, and research data into actionable insights at an unprecedented scale, effectively compressing months of analysis into hours.

Combining high-performance computing with AI enables the project to achieve its mission of transforming how federal agencies operate, supporting missions across national security, scientific research, healthcare, and industrial innovation. 

For the government, this investment by Amazon will boost speed, security, and analytical capacity, strengthening America’s leverage in AI and technology.

One of its strategic partners is Anthropic, the startup now sitting on a $13.8 billion investment from Amazon.

TechCrunch confirmed through two sources directly familiar with the development that this is a calculated strike at a fragmented and increasingly competitive sector. 

AWS’ offering will give startups a central hub to publish their AI agents while giving enterprise users a one-stop shop to search, test, and deploy the tools they need.

This marketplace will offer plug-and-play integration, allowing customers to deploy agents directly within AWS environments. Startups, in turn, will be able to charge fees based on usage or subscriptions, an approach similar to how Software-as-a-Service (SaaS) models operate. 

AWS will take a revenue cut, but insiders say it’s deliberately minimal to entice developers.

Notably, Google Cloud launched its Agent Marketplace back in April. Microsoft followed suit in May with its Agent Store inside Microsoft 365 Copilot. Salesforce and ServiceNow also offer similar platforms for enterprise workflows. Amazon’s late entry may raise eyebrows, but its execution appears bigger in scope and more commercially aggressive.

What gives AWS a potential edge is infrastructure. With its own cloud powering the backend, AWS can offer developers access to hundreds of thousands of GPUs for building, training, and running agents, something most competitors simply can’t match.

Anthropic’s involvement, meanwhile, may prove critical. The company’s Claude model, already considered one of the few credible rivals to OpenAI’s GPT-40, will form the backbone of many of its own agentic offerings. 

Anthropic also enables third-party developers to build agents using its API, and with its revenue reportedly hitting $3 billion annually as of May 2025, its model appears to be resonating with the enterprise crowd.

“The marketplace could help us reach a broader set of users, especially those who are already operating in AWS environments or looking for interoperability,” one Anthropic insider said, requesting anonymity due to lack of public authorisation.

Essentially, this launch also aims to solve what’s fast becoming one of the most annoying problems in the AI space: fragmentation. Today, most AI agents live inside walled gardens. 

A customer who uses agents built on one cloud can’t easily deploy or integrate them into another system. By unifying offerings under one marketplace, AWS hopes to simplify that chaos, and grab a major piece of what analysts expect to become a $50.3 billion market by 2030.

Even with this, everyone is wondering if these marketplaces actually bring value to smaller startups or get swallowed by dominant players. That outcome may depend less on marketplace design and more on execution, and whether AWS can place itself not just as a provider, but as the centre of gravity in the emerging agent economy.

In the meantime, AWS is betting heavily that developers want more than just raw compute, they want access, scale, and visibility. With Anthropic on board and infrastructure already in place, Amazon may be late to the party, but it didn’t come unarmed.