The World Economic Forum (WEF) and Amazon logos are displayed during the 56th World Economic Forum (WEF) Annual Meeting in Davos, Switzerland, January 20, 2026.
Romina Amato | Reuters
Amazon CNBC confirmed at Tuesday’s Retail Technology Conference that it plans to address a series of recent failures, including one related to AI-assisted coding errors.
Dave Treadwell, the top executive who oversees the technology infrastructure for Amazon’s websites, told employees that the company’s “This Week in Store Tech” (TWiST) conference will be a “deep dive” into “some of the issues that got us here.” The meeting is scheduled to begin at 12:30 PM ET.
“As you are all aware, the availability of our site and related infrastructure has been poor lately,” Treadwell, the eCommerce Foundation’s senior vice president, wrote in a memo to employees seen by CNBC. He added that the focus of the meeting was changing “to account for Severity 1 occurrences,” referring to high-severity incidents that cause critical system outages or performance degradation.
Treadwell said Amazon experienced four such incidents in one week, noting that a thorough investigation is needed to “restore a strong availability posture.”
The Financial Times first reported on the memo. TWiST is a weekly meeting in which retail technology leaders review the performance of store operations, an Amazon spokesperson said.
“As part of our normal business operations, the meeting will also include a review of the availability of our website and apps as we focus on continuous improvement,” a spokesperson said in a statement.
The meeting comes after Amazon’s online store experienced outages for some users last week. For about six hours on Thursday, users of the website and app were unable to check out, access account information or view product prices. Amazon said in a statement that the issue was related to “software code deployment.”
Amazon and its hyperscaler rivals are increasing spending on infrastructure to cope with a surge in demand for artificial intelligence services that require increasing computing power. Amazon said in its earnings call last month that it expects to spend $200 billion in capital expenditures this year, more than its peers.
At the same time as it increases spending on AI, Amazon continues to cut jobs. The company implemented large-scale layoffs in October, cutting about 14,000 positions, and then laid off about 16,000 employees in January. Amazon also laid off more than 27,000 employees from 2022 to 2023.
In a separate memo to staff, Treadwell wrote that “genAI-powered changes” contributed to recent incidents dating back to the third quarter of 2025.
Among other factors, he pointed to “GenAI tools supplementing or accelerating production change orders, leading to unsafe practices,” according to a memo seen by CNBC. Treadwell also acknowledged that “best practices and safeguards” around the use of generative AI are not yet fully established.
The memo said Amazon plans to “enhance” various safeguards to prevent further problems, including requiring more senior engineers to review “GenAI-assisted” production changes made by lower-level staff.
“We are implementing temporary safeguards that introduce controlled friction into changes to the most important parts of the retail experience. In parallel, we are investing in more durable solutions that include both deterministic and agentic safeguards,” Treadwell wrote.
Amazon Web Services has also suffered several outages in recent months, but the company announced Tuesday that its cloud group was not involved in the incidents mentioned by Treadwell.
According to multiple reports, AWS suffered an incident in December that caused its cost management functionality to go out for an extended period of time. The FT reported that the problem arose after allowing engineers to make changes to the company’s Kiro AI coding tool.
Amazon said in a statement at the time that the outage was due to “user error” and not AI.
Beware: Amazon is an undervalued stock with opportunities.
