On Friday Eastern Time, 20 leading technology companies including Microsoft, Meta, Google, and Amazon jointly announced at the 60th Munich Security Conference that they will work together to combat erroneous information generated by artificial intelligence technology in this year's elections.
20 technology companies jointly sign an agreement
This year is a global election year, and multiple countries including India and the United States will hold elections. According to statistics, this year's elections will affect over 4 billion people in more than 40 countries. Artificial intelligence technology, especially DeepFake and other technologies, can produce deceptive audio, video, and images to mimic key stakeholders in democratic elections, or provide false voting information, thereby threatening the fairness of elections.
In order to reduce this harm, 20 technology companies have jointly signed an agreement. Among the companies participating in this agreement, there are technology giants such as Microsoft, Meta, Google, Amazon, IBM, Adobe, and chip design company Arm, as well as AI startups such as OpenAI, Aerospace, and Stability AI, as well as social media companies such as Snap, TikTok, and X.
The companies participating in the agreement have agreed to eight commitments, including assessing model risks, seeking to detect and address the dissemination of such false election content on their platforms, and providing transparency to the public in these processes.
Like most voluntary commitments in the technology and other industries, the press release clearly states that these commitments only apply to "areas related to the services provided by each company.".
"Democracy relies on secure and reliable elections," Google's Global Affairs President Kent Walker said in this press release, reflecting the industry's efforts to address "election misinformation generated by artificial intelligence that erodes trust.".
Christina Montgomery, IBM's Chief Privacy and Trust Officer, stated in a press release that in this critical election year, "specific collaborative measures need to be taken to protect people and society from the amplification risk of deceptive content generated by artificial intelligence."
The interference of AI on elections is causing concern
According to data from machine learning company Clarity, the rise of AI generated content has raised concerns about serious election related misinformation, particularly with the number of DeepFake content created this year increasing by 900% year-on-year.
False information in elections has always been a major issue, dating back to 2016. Nowadays, with the rapid rise of artificial intelligence, governments around the world are paying more attention to this issue.
"We have reason to be seriously concerned that artificial intelligence may be used to mislead voters in elections," California Democratic Senator Josh Becker said in an interview. "It's encouraging to see some companies coming to the negotiating table, but I haven't seen enough details yet, so we may need legislation to establish clear standards."
However, technology companies still have a long way to go to truly and effectively address the impact of artificial intelligence on elections. For example, detection and watermarking techniques for identifying DeepFakes have not yet developed fast enough. At present, various companies have only reached an agreement on a set of technical standards and testing mechanisms.