Twenty major technology companies, including Google, Meta, Microsoft, OpenAI, TikTok, X, Amazon and Adobe, vowed Friday to prevent deceptive uses of artificial intelligence from interfering with global elections.
In 2024, 4 billion people in more than 40 countries are expected to vote. Experts said easily accessible generative AI tools could influence elections and sway votes if they fell into the wrong hands.
Generative AI tools allow users to create images, videos, or audio from simple text prompts. Some don’t have the necessary safeguards in place to prevent users from creating content in which politicians or celebrities say things they’ve never said or done things they’ve never done.
Just last month, a fake robocall pretending to be President Joe Biden’s voice discouraged individuals from voting in the New Hampshire primary. In Taiwan, a significant amount of deepfakes were circulated on social media ahead of the January 13 presidential election.
At the Munich Security Conference, these companies announced a “Technology Agreement to Combat Deceptive Use of AI in the 2024 Elections.” This is a voluntary agreement containing eight specific commitments to deploy technologies to counter harmful AI content.
“Democracy is based on safe and secure elections,” Kent Walker, Google’s president of international affairs, said at a conference in Munich. “Google has supported election integrity for years, and today’s agreement reflects our industry’s commitment to combating AI-generated election misinformation that undermines trust. Don’t let digital abuse threaten the opportunities that improve our economy, create new jobs, and pass on AI for generations to come.”Advancing health and scientific advances. ”
These efforts include detecting the distribution of this content on our platforms and developing and implementing technology to reduce the risks associated with deceptive AI content.
“We are pleased that these companies are taking steps to voluntarily curb what we believe to be the worst forms of abusive behavior,” said Lisa Gilbert, executive vice president at Public Citizen. will help, but it will not be enough.” The nonprofit organization has been advocating for political and blatant deepfake legislation, it said in a press release.
“AI companies are committed to suppressing technologies that pose significant election risks, especially text-to-video conversion, until substantial and sufficient safeguards are in place to help avoid many potential problems.” We have to,” she added.
This follows Open AI’s promotion of its new text-to-video conversion tool, Sora, which is not yet publicly available and has been tested by safety experts for potential risks. That’s where I am.
“This is a constructive step forward. We thank the companies involved for their proactive approach. Time will tell how effective these measures are and whether further action is needed. We all They have a role to play in protecting election integrity. We support the democratic process and welcome this important step by technology companies.” Sen. Graham, R-South Carolina, said in a joint statement in support of the agreement.
Signatories of the agreement as of today include Adobe, Amazon, Anthropic, Arm, Eleven Labs, Google, IBM, Inflection AI, LinkedIn, McAfee, Meta, Microsoft, Nota, OpenAI, Snap Inc., Stability AI, TikTok, and Trend is. Micro, Truepic, X.
The companies said they encourage governments to take meaningful steps to legislate against the harmful uses of AI and look forward to more companies joining.