‘Deepfake’ of Biden’s Voice Called Early Example of US Election Disinforma

　　‘Deepfake’ of Biden’s Voice Called Early Example of US Election Disinformation

　　被称为美国选举虚假信息的早期例子——拜登的“深度伪造”语音

---------------------------------------------------------------------------------

　　By Bryan Lynn

　　24 January 2024

　　A fake voice message claiming to be U.S. President Joe Biden was recently sent to voters in an example of possible election disinformation.

　　The voice message is an example of what is known as a "deepfake." A "deepfake" is a piece of audio or video created to make it appear that people in it are saying or doing things that they never did. Newly developed artificial intelligence (AI) tools make it easier for people to make and publish such content.

　　The fake voice of Biden was included in a political message sent by telephone directly to voters in the northeastern state of New Hampshire. This marketing method, known as "robocalling," involves a machine that calls large numbers of people to play them a prerecorded message.

　　The calls are being investigated by New Hampshire election officials. The state recently held America's first presidential primary contest. Primary elections in the United States usually choose the candidates for the presidential election in November.

　　It is not known who created the fake audio, which was heard by reporters from several news organizations. In it, Biden seems to try to persuade voters not to take part in the primary.

　　The voice says, "Voting this Tuesday only enables the Republicans in their quest to elect Donald Trump again." The message continues, "Your vote makes a difference in November, not this Tuesday."

　　Voting in state primaries does not affect a person's vote later in the U.S. presidential election, which is set for November 5, 2024.

　　The Biden robocalls even included language the U.S. president has often used in the past, specifically the phrase, "What a bunch of malarkey." The phrase is used to describe something foolish, unreasonable or meaningless. The phone message advises listeners to "save your vote for the November election."

　　It is not known how many people received the call with Biden's voice. Biden did not campaign in New Hampshire and his name was not on the state's primary ballot. This is because his campaign decided to begin the primary process in the state of South Carolina instead. However, a New Hampshire state law requires it to hold the nation's first primary.

　　White House press secretary Karine Jean-Pierre confirmed the call was "fake and not recorded by the president." The campaign for former President Donald Trump said it was not involved with the call.

　　The head of Biden's re-election campaign, Julie Chavez Rodriguez, said in a statement the organization was investigating the incident and "actively discussing additional actions to take." Rodriguez said the robocall was a clear attempt "to suppress voting and deliberately undermine free and fair elections."

　　Robert Weissman is head of the nonprofit citizen activist group Public Citizen. He said in a statement the Biden robocall provided fresh evidence that "the political deepfake moment is here." He urged governments to consider passing legislation to limit the use of deepfakes to prevent election "chaos."

　　Deepfakes created with AI technology have already appeared in campaign advertisements in the 2024 presidential race, the Associated Press (AP) reports. The technology has also been used to spread false information during election campaigns across the world, from Slovakia to Indonesia to Taiwan.

　　Hany Farid is a digital investigations expert at the University of California, Berkeley. He told the AP the Biden robocall demonstrates how AI methods can be "weaponized" in elections. Farid added that recent cases of audio and video used to mislead voters "is surely a sign of things to come."

　　As AI technologies improve, governments around the world are seeking ways to restrict them from causing public harm. The U.S. Congress has yet to pass legislation to limit the use of such technologies in the political process. The Federal Election Commission is currently considering public comments on a petition for it to restrict AI-created deepfakes in campaign advertisements.

　　David Becker heads the nonprofit Center for Election Innovation and Research. He told the AP the use of deepfakes is just the latest tool in a long history of "dirty tricks" aimed at influencing the political process.

　　Becker added that the goal of such efforts can be to confuse voters to a point where they no longer recognize false information from the truth. "They don't need to convince us that what they're saying, the lies they're telling us, are true," he said. "They just need to convince us that there is no truth, that you can't believe anything you're told."

　　I'm Bryan Lynn.

　　Bryan Lynn wrote this story for VOA Learning English, based on reports from The Associated Press, Reuters and online sources.

---------------------------------------------------------------------------------

　　作者：Bryan Lynn

　　2024年1月24日

　　最近，一条声称是美国总统乔·拜登的假语音信息被发送给选民，这是可能的选举虚假信息的一个例子。

　　这条语音信息是所谓的“深度伪造”的一个例子。"深度伪造"是一种音频或视频，其创建的目的是使其中的人看起来在说或做他们从未做过的事情。新开发的人工智能（AI）工具使人们更容易制作和发布此类内容。

　　拜登的假语音被包含在一个直接发送给美国东北部新罕布什尔州选民的政治信息中。这种营销方法，被称为"机器人电话营销"，涉及一台机器，它给大量的人打电话，播放他们预先录制的信息。

　　新罕布什尔州的选举官员正在调查这些电话。该州最近举行了美国的首次总统初选比赛。在美国，初选通常选出11月总统选举的候选人。

　　目前还不清楚是谁制作了这段假音频，这段音频已经被几家新闻机构的记者听到。在音频中，拜登似乎试图劝说选民不参加初选。

　　语音说：“本周二投票只会助长共和党再次选举唐纳德·特朗普的企图。”信息继续说：“你的投票在11月会有影响，而不是本周二。”

　　在州初选中投票不会影响一个人在2024年11月5日设定的美国总统选举中的投票。

　　拜登的机器人电话甚至包括了美国总统过去经常使用的语言，特别是“一堆胡说八道”的短语。这个短语用来描述一些愚蠢的、不合理的或无意义的事情。电话信息建议听众“为11月的选举保存你的投票”。

　　目前还不清楚有多少人接到了带有拜登声音的电话。拜登并没有在新罕布什尔州进行竞选活动，他的名字也没有出现在该州的初选选票上。这是因为他的竞选团队决定在南卡罗来纳州开始初选过程。然而，新罕布什尔州的一项法律要求它举行全国的第一次初选。

　　白宫新闻秘书卡琳·让-皮埃尔确认电话是“假的，不是总统录制的”。前总统唐纳德·特朗普的竞选团队表示，他们与这个电话无关。

　　拜登的连任竞选团队负责人朱莉·查韦斯·罗德里格斯在一份声明中表示，该组织正在调查这起事件，并“积极讨论采取额外的行动”。罗德里格斯说，这个机器人电话明显是“压制投票和故意破坏自由公正选举”的尝试。

　　非营利公民活动组织公民公众的负责人罗伯特·魏斯曼在一份声明中说，拜登的机器人电话提供了新的证据，表明“政治深度伪造的时刻已经到来”。他敦促政府考虑通过立法限制深度伪造的使用，以防止选举“混乱”。

　　据美联社报道，已经有使用AI技术创建的深度伪造出现在2024年总统竞选的广告中。这项技术也已经被用来在全球范围内的选举活动中传播虚假信息，从斯洛伐克到印度尼西亚再到台湾。

　　加利福尼亚大学伯克利分校的数字调查专家Hany Farid告诉美联社，拜登的机器人电话展示了AI方法如何在选举中被“武器化”。Farid补充说，最近用来误导选民的音频和视频的案例“肯定是未来事情的预兆”。

　　随着AI技术的改进，世界各地的政府正在寻找方法来防止它们对公众造成伤害。美国国会尚未通过立法限制在政治过程中使用这种技术。联邦选举委员会目前正在考虑公众对一项要求其限制在竞选广告中使用AI创建的深度伪造的请愿的评论。

　　非营利性选举创新与研究中心的负责人David Becker告诉美联社，使用深度伪造只是长期以来旨在影响政治过程的“肮脏手段”的最新工具。

　　Becker补充说，这种努力的目标可能是让选民混淆到他们无法再从真相中辨认出虚假信息的地步。他说：“他们不需要说服我们，他们告诉我们的谎言是真的。他们只需要说服我们，没有真相，你不能相信你被告知的任何事情。”

　　我是Bryan Lynn。

　　Bryan Lynn为VOA学习英语撰写了这篇报道，报道基于美联社、路透社和在线来源的报道。

---------------------------------------------------------------------------------

　　这个故事中的词汇

　　fake – adj. 假的，不真实的

　　quest – n.尝试获取或做某事困难的行为

　　deliberate – adj. 故意的，或计划的

　　undermine – v. 削弱某事物前进的有效性或能力

　　chaos – n. 一种完全没有秩序的情况

　　petition – n. 正式请求做某事的许可

　　confuse – v. 使某人无法清晰地思考或理解某事

　　convince – v. 使某人相信某事是真实的