AWS GenAI models safe from harmful, bias content: Swami Sivasubramanian

Las Vegas: Amazon is taking all the necessary steps to make sure that its generative AI models are constantly monitored for any toxicity and bias so that future AI models can be built in a safe and responsible manner, says Swami Sivasubramanian, Vice President of Database, Analytics, and Machine Learning at Amazon Web Services (AWS).

As governments, including in India, gear up to tame fake AI-generated images and videos (deepfakes), Sivasubramanian told IANS that AWS will continue to invest more and more in order to continue promoting the safe, secure, and transparent development of AI technology that benefits society.

“We take these issues super seriously and have been engaging with policymakers on developing responsible AI. We were one of the few companies which made voluntary commitments in the White House on how we build these AI models, especially the future models, in a safe and responsible manner,” Sivasubramanian emphasised.

At AWS, “We are committed to continued collaboration with the policymakers, the technology industry, researchers, and the AI community to advance the responsible and secure use of AI,” he added.

Amazon’s Titan family of foundation models is built to detect and remove harmful content in the data that customers provide for customisation, reject inappropriate content in the user input, and filter the model’s outputs containing inappropriate content such as hate speech, profanity, and violence.

Amazon CodeWhisperer is the only AI coding companion with built-in security scanning for finding and suggesting remediations for hard-to-detect vulnerabilities.

“Amazon SageMaker Clarify detects and measures potential bias using a variety of metrics so developers can address potential bias and explain model predictions,” said Sivasubramanian who has been working in Amazon for more than 15 years.

At its flagship ‘re: Invent 2023’ conference here, AWS launched Amazon Titan Image Generator that can create new images with text description or customise existing images.

As a content creator, you can now use the image generator to quickly create and refine images using English natural language prompts.

“You can use the model to easily swap out an existing background to the background of a rainforest, or can use the model to seamlessly swap out backgrounds to generate lifestyle images, all while retaining the main subject of the image and to create a few more options,” said Sivasubramanian.

Titan Image Generator also mitigates harmful content generation to support the responsible use of AI.

“All images generated by Amazon Titan contain an invisible watermark, by default, designed to help reduce the spread of misinformation by providing a discreet mechanism to identify AI-generated images,” Sivasubramanian told IANS.

The image generator will help companies in advertising, e-commerce, and media and entertainment to create studio-quality, realistic images in large volumes and at low cost.

In July, Amazon joined US President Joe Biden and leaders across government and industry to commit to continue promoting the safe, secure, and transparent development of AI technology that benefits society.

“We build AI with responsibility in mind at each stage of our comprehensive development process. Throughout design, development, deployment, and operations we consider a range of factors including accuracy, fairness, appropriate usage, toxicity, security, safety, and privacy,” Sivasubramanian noted.

(Nishant Arora can be reached at nishant.a@ians.in)

–IANS

Comments are closed.