AWS AI service cards signal Amazon’s responsible AI catch-up rate

Amazon launched at AWS re:Invent this week AWS AI service Cardsa form of responsible AI documentation intended to help customers better understand the AI ​​services offered by the cloud computing leader.

According to Gartner analyst Svetlana Sicular, the AWS AI Service Cards are a signal that Amazon is taking steps to catch up with its competitors in embracing responsible AI.

In the past, Amazon “denied responsible AI” and lagged behind its competitors, including Microsoft and Google, in addressing responsible AI issues, Sicular told VentureBeat.

But Amazon has “a very good history of catching up when they put their minds and effort and resources into catching up,” she said. “They realized this is serious and in everyone’s interest to have transparency, because otherwise they are not competitive.”

AWS AI service cards provide a responsible AI repository

According to an Amazon blogging after, the AWS AI Service Maps “provide customers with a single place to find information about intended use cases and constraints, responsible AI design choices, and deployment and performance optimization best practices for our AI services.” The company said it addresses issues such as fairness and bias, explainability, robustness, governance, transparency, privacy and security.

The first three available service cards are for software dealing with sensitive demographic issues, including Amazon T extract – AnalysisID, Amazon Transcribe – Batch (English-US).and Amazon Recognition – Face Matching. The third card addresses issues related to the controversial tool that made headlines in 2019 when Amazon challenged a study that said the technology struggled to identify the gender of dark-skinned individuals. Following the 2020 murder of unarmed black man George Floyd during an arrest, the company issued a moratorium on police use of its facial recognition software.

Now the new AWS AI Service Card says it doesn’t support matching “images that are too blurry and grainy for a human to recognize the face, or where large areas of the face are occluded by hair, hands, and other objects. ”

Amazon’s recent steps toward responsible AI

Sicular pointed out that Amazon has taken steps toward responsible AI in recent years, including hiring the authors of The ethical algorithmAaron Roth and Michael Kearns, as Amazon Scholars in 2020.

Kearns, a professor at the University of Pennsylvania — and an Amazon scientist since 2020 — told Reuters that the decision to release the cards followed privacy and fairness audits of the company’s software. He said the cards would publicly address ethical concerns about AI, with technical regulation on the horizon.

“The most important thing about this launch is the commitment to doing this continuously and comprehensively,” he said.

The AWS AI Service Cards build on related efforts from Microsoft, Google and IBM. “The launch of the AWS AI Service Cards is a step in the right direction toward building responsible AI by design,” said Krishnaram Kenthapadi, chief scientist at ML observatory firm Fiddler AI. “This builds on a related line of work such as datasheets, model charts, factsheets, system charts and ABOUT ML towards benchmarking and transparency of datasets, models and systems for machine learning,” he said.

“They are very good first steps,” said Sicular, adding that she thinks Amazon is “likely to move up the ranks” in terms of reaching executive decision makers at a time when there is fierce competition among the major cloud providers for AI investment. in the cloud.

“I think it’s part of the competition with Microsoft,” she explained. “But [Amazon] places great emphasis on responsibility [now]. That is very impressive.”