Google Jarvis AI Extension Leaked On Chrome Store

Alphabet’s Google is at the centre of what appears to be an accidental leak of its upcoming AI agent, known as Jarvis AI.

According to the Information, Google reportedly posted on its Chrome extension store, an internal preview of an upcoming AI project, codenamed Jarvis (apparently named after the AI character that first featured in Marvel’s Ironman films).

Jarvis AI reportedly will browse the web for a user in Chrome, but Google quickly removed the AI agent listing from the Chrome store, once its mistake was realised.

Jarvis AI

According to the report, Google’s listing for Jarvis AI, had stated it is “a helpful companion that surfs the web with you.”

But it seems that despite Google quickly pulling the extension listing, some users were able to download the Jarvis AI prototype before it was pulled.

However those that did download it were reportedly unable to use it, due to the program reportedly requiring specific access permissions.

The Information reported that Google plans to officially unveil Jarvis in December.

So what exactly does Jarvis AI actually do?

Well according to the report, Jarvis AI is expected to allow users to automate certain web browsing tasks from within Google Chrome.

This will allow the user to concentrate on other things while the AI agent shops or even books flights for example.

AI agents

Google is not the only tech firm thought to be developing AI agents, which are essentially independent artificial intelligence models capable of performing a range of tasks without human input.

Last month Anthropic (which is currently seeking more funding) released a product similar to Jarvis AI, namely the experimental Claude 3.5 Sonnet.

Anthropic said in October that “available today on the API, developers can direct Claude to use computers the way people do – by looking at a screen, moving a cursor, clicking buttons, and typing text. Claude 3.5 Sonnet is the first frontier AI model to offer computer use in public beta.”

OpenAI meanwhile has already confirmed it will allow developers to build their own OpenAI Agent, with its ‘agents’ expected to be officially launched next year.

That is not the only recent development from OpenAI, after it last week began rolling out search engine features that it had begun testing in July, putting the AI start-up in more direct competition with Google and major stakeholder Microsoft.

Tom Jowitt

Tom Jowitt is a leading British tech freelancer and long standing contributor to Silicon UK. He is also a bit of a Lord of the Rings nut...

Recent Posts

Amazon Mulls New Multi-Billion Dollar Investment In Anthropic – Report

Amazon is reportedly in talks to pump billions of dollars more into AI start-up Anthropic,…

7 hours ago

FTX’s Caroline Ellison Begins Her Two Year Prison Sentence

Star witness for the US prosecution of FTX founder Sam Bankman-Fried, has begun her two…

7 hours ago

More Layoffs For iRobot Staff After Abandoned Amazon Deal

After axing 31 percent of its workforce when it failed to be acquired by Amazon,…

1 day ago

Mozilla Foundation Confirms Layoffs, Eliminates Advocacy Division

Mozilla Foundation axes 30 percent of its staff, and is eliminating its Advocacy Division that…

1 day ago

Google To Make MFA Mandatory Next Year

Improving security. Mandatory multi-factor authentication (MFA) is coming to the Google Cloud by the end…

1 day ago

UK Government Launch AI Safety Platform For Businesses

New AI assurance platform from UK government will help businesses ensure they can safely develop…

1 day ago