In a recent speech in Washington (USA), Mr. Smith said his biggest concern about AI is deepfake, a technology that creates images and videos that look realistic. He called for measures to help people distinguish between real and fake photos and videos.
“We have to address the issues around deepfakes,” the Microsoft president said. “We have to take steps to combat the use of AI to alter legitimate content to mislead or deceive people.”
He also called for licensing the most important forms of AI with “obligations to protect safety, physical security, cybersecurity, national security.” We need new export controls, at least a step up from existing ones, to ensure that AI models are not stolen or used in ways that violate export controls, he said.
AI regulation has been a hot topic since the public launch of OpenAI’s ChatGPT chatbot. Washington lawmakers are grappling with how to introduce AI regulation as companies large and small rush to bring AI to market.
Last week, Sam Altman, CEO of OpenAI, told the US Senate that using AI to interfere with election integrity was a “big concern.” He also argued that AI needs to be regulated, calling for global cooperation in this area.
Smith argued that people need to be held accountable for any problems caused by AI. He urged lawmakers to put “safety brakes” on AI technologies used to manage power grids, water supplies, and other critical infrastructure so that humans can take control. He also mentioned a “Know Your Customer” (KYC)-like system for developers of powerful AI models to know how their technology is being used, and to inform the public about what AI-generated content is being used to identify fake videos.
Earlier this week, an AI image of an explosion near the Pentagon went viral on social media and was shared by many “blue-ticked” accounts. The source of the image is unclear, and although the US Department of Defense confirmed that it was a fake, it still caused the stock market to drop significantly within minutes.
OSINTdefender , a blue-ticked Twitter account with more than 336,000 followers, was one of the sites that shared the photo. The account owner later apologized for spreading false information, saying it was an example of how “images like this can be used to manipulate information with ease.”
It's also one of several popular AI images in recent weeks, including the Pope wearing a stylish white coat, the black-and-white photo that won the Sony World Photography Award, and former President Donald Trump being arrested by police.
(According to CNN, The Guardian)
Source
Comment (0)