A consortium of 20 prominent technology firms made a collective pledge on Friday to address the issue of AI misinformation in the lead-up to the 2024 elections. The focus of this initiative is specifically on combating deepfakes, which utilize misleading audio, video, and images to replicate influential figures in democratic elections or disseminate false voting details.
Among the signatories of this agreement are Microsoft, Meta, Google, Amazon, IBM, Adobe, and chip manufacturer Arm. Additionally, emerging artificial intelligence startups like OpenAI, Anthropic, and Stability AI have joined forces with social media giants such as Snap, TikTok, and X.
With numerous significant elections on the horizon, impacting over four billion individuals across more than 40 nations, tech platforms are gearing up to address the escalating concerns surrounding election-related misinformation. The proliferation of AI-generated content has exacerbated these worries, with the volume of deepfakes created witnessing a staggering 900% surge year over year, as reported by Clarity, a machine learning company.
Despite this surge, the progress in developing detection and watermarking technologies to identify deepfakes has not kept pace with the rapid evolution of such deceptive content.
This announcement follows closely on the heels of OpenAI introducing Sora, its latest AI-generated video model, which functions akin to the organization’s image-generation tool, DALL-E. Sora enables users to input a scene description, prompting the generation of a high-definition video clip. Moreover, it can produce video clips inspired by static images, extend existing videos, or fill in missing frames.
The participating companies have committed to eight overarching pledges, including evaluating model risks, proactively addressing the dissemination of such content on their platforms, and ensuring transparency regarding these processes for public scrutiny. Notably, the scope of these commitments is delineated to apply solely to services provided by each respective company.
Kent Walker, Google’s president of global affairs, emphasized the critical role of secure and trustworthy elections in upholding democracy. He underscored the industry’s resolve to confront the menace of “AI-generated election misinformation,” which undermines public confidence in the electoral process.
Echoing this sentiment, Christina Montgomery, IBM’s chief privacy and trust officer, stressed the imperative for tangible collaborative actions in this pivotal election year to shield individuals and societies from the heightened risks posed by AI-generated deceptive content.
