AWS bets big on Amazon Q to take on ChatGPT, Microsoft Copilot

AWS’ bid to take on ChatGPT and Microsoft’s Copilot is an enterprise AI tool that can build apps, write code and take on tasks like a junior worker would.

The company is supremely confident of Amazon Q’s ability to deliver accurate information. Picture: Reuters

AWS has just launched what it claims to be the most powerful enterprise AI tool in the market, combining some of the key elements of ChatGPT and Microsoft’s Copilot.

The US tech behemoth is pushing to gain a significant market share, having developed an AI tool that can work with up to 40 different vendors and applications, including Gmail, Salesforce and Zendesk.

It’s called Amazon Q, and one of the key AWS executives behind it was not shy to share that there was some “underlying technology similarity” with key players already in the AI space.

AWS director of AI developer experiences Doug Seven said one of the big differences with Amazon Q was that it better respected the privacy of users.

“What is really different about Amazon Q is the work we’ve done in the category of responsible AI and what we’ve done around security to ensure that Q is respecting all of your security permissions and boundaries,” Mr Seven said.

Amazon Q would not share specific data with users if they weren’t given access by the employer, even if the tool could bypass those restrictions.

The process of setting up privacy permissions for Amazon Q could be done in a matter of hours, Mr Seven said.

AWS director of AI developer experiences Doug Seven.

Amazon Q can build apps, search a business’s entire database to answer queries, write new code and improve existing code, detect security flaws and be assigned tasks.

The tool is part of AWS’s bid to gain an even bigger slice of the enterprise market which Google also recently made a major play into.

Google last month released a suite of new additions to its Gemini models, including the ability to strengthen cybersecurity, edit videos – including creating voiceovers – as well as slash medical diagnosis turnaround times and analyse company earnings reports.

Google has been accused of being slow to the AI race but conversely AWS has been on the front foot with public-facing AI products long before the launch of ChatGPT.

Mr Seven indicated that there was some hope that the launch would steer companies away from believing that every expert in their respective field had to become an AI expert as well.

“It’s really around this idea that being a subject matter expert means you should also have to be a generative AI expert or a technology expert to build an application,” he said.

“They should be able to just express themselves in natural language in Amazon Q and it can build that application for them.”

Mr Seven was so confident in Amazon Q’s ability to deliver accurate information that he said if a user were getting the wrong information they should check how they asked the program.

If a company were to cease using Amazon Q, they would retain all of their data and code.

“Any code that is generated by Amazon Q belongs to the customer. It’s all based on their context and their input and it’s generated on their behalf and so it’s their code,” he said.

On the ability to code, about 50 per cent of Amazon Q’s suggestions had been approved by NAB, according to chief engineer Andrew Brydon, who said it delivered “unparalleled benefits”.

“So far, our developers have accepted 50 per cent of the code suggestions made by Amazon Q Developer, and that number continues to increase as we scale,” he said.