Amazon has unveiled a new suite of artificial intelligence tools designed to transform how companies recruit workers, manage operations, and automate decision-making, marking a significant push into “agentic” AI systems that can operate with minimal human oversight.
The Seattle-based technology giant, which hires hundreds of thousands of seasonal employees each year for its holiday operations, on Tuesday introduced new software aimed at speeding up recruitment by reducing the need for traditional, face-to-face interviews. Alongside this, the company outlined a new AI design philosophy it calls “humorphism”, which it says is intended to make artificial intelligence more human-centred.
Amazon described “humorphism” as a framework that “humanizes AI and adapts to how humans work, not the other way around.”
The announcements were made at a company event expected to feature Amazon Web Services CEO Matt Garman and executives from OpenAI, underscoring the increasingly competitive race among major technology firms to define the future of enterprise AI.
The launch comes amid Amazon’s expanding investment in artificial intelligence infrastructure. In February, the company said it would invest up to $50 billion in OpenAI, while Microsoft recently signalled it would lose exclusive access to some OpenAI technology, allowing the ChatGPT maker to broaden its commercial partnerships.
A central focus of the event is the development of autonomous AI systems known as “agents”, which are designed to execute tasks, make decisions, and manage processes with little to no human input. While proponents argue these systems could dramatically improve efficiency, they have also raised concerns around oversight, safety, and potential job displacement.
Other major technology firms are moving in a similar direction. Alphabet recently indicated it is also expanding its enterprise AI strategy with agent-based systems, following similar efforts from OpenAI and Anthropic.
At the centre of Amazon’s workplace automation push is a new hiring platform called Connect Talent. The software is designed to help companies recruit at scale, particularly during peak periods such as the holiday retail season, when demand for temporary staff surges.
Connect Talent uses artificial intelligence to conduct interviews around the clock, screen candidates, and generate recruitment notes for hiring teams without human intervention. Amazon itself hired around 250,000 seasonal workers ahead of last year’s holiday period.
Colleen Aubrey, Senior Vice President of Applied AI Solutions at AWS, said candidates would be made aware when AI is being used in the hiring process, adding that the system is still being refined to make it sound more natural.
“The experience continues to get better and better each iteration we go through,” she said in a briefing with Reuters before the event. “There’s some art around making that voice interaction natural and human.”
Aubrey also explained the broader thinking behind Amazon’s “humorphism” approach, saying it reflects an effort to embed human behaviours into AI systems.
“How do we translate the human behaviors of working together into a product?” she said. “That’s what we’re going after and hopefully you’ll see that.”
She added that the philosophy is intended to ensure AI tools reflect real-world workplace dynamics, even as concerns grow about automation reducing the need for human labour. Amazon has already linked some of the roughly 30,000 corporate job cuts it has made since October to efficiency gains driven by AI.
Beyond recruitment, Amazon also introduced a new tool called Connect Decisions, designed to support supply chain planning and procurement. The system uses AI to analyse and compile large volumes of operational data, helping businesses make faster and more informed decisions.
Aubrey said Amazon’s own logistics experience helped shape the product.
With Connect Decisions, she said, companies will be “able to have AI do that work behind the scenes and be able to equip a planner with the data that they need.”
Boluwatife Enome
Follow us on:
