Polish Opposition Condemned for Using AI-Generated Deepfake

    Published on:

    Poland’s largest opposition party, Platform for the Citizens of the Centrist (PO), has faced heavy criticism for using clips of the prime minister’s speech manipulated by artificial intelligence in an election campaign ad.

    The 2023 Polish parliamentary elections are scheduled to take place on October 15 to elect members of the Sejm and Senate.

    “Guided by the positive comments I have just received from the National Electoral Commission regarding the proposed dates for the Seym and Senate elections, I have decided to order these elections on October 15, 2023. ,” the document reads. translation Election day announcement by President Andrzej Duda.

    Also read: Could AI influence the next presidential election?

    As the election approaches, PO has been heavily criticized for using AI-driven ads. Party representatives were forced to admit and clarify that additional content had been added to the content of the advertisement.

    Extreme criticism of PO

    “Ordinary deepfakes. Campaigns on this platform do not contain anything real.” answered Polska Swerenna (Polish Sovereignty), a Catholic nationalist party, tweeted PO.

    “And where in the film or video is it shown that the sound was generated by AI?” Magog questioned.

    Mateusz Morawicki’s voice is produced in such a way that the recipient will immediately recognize it as the produced speech. Most of all, even a child knows that Morawitzki didn’t read his emails to Dvorczyk aloud, but this synthesized speech needs to be clearly marked.

    In addition to the criticism, some Twitter users said that people who watch such manipulated videos outside of X would not read the comments and try to understand whether it was real or fake, leading to the issue of fake misrepresentations. Some people are wary.

    “Most people will be watching this video provided by someone else, even from outside X, so it is not possible to read that the comments were generated by AI. It sets a precedent for tampering with statements by politicians, even as part of .Things like this should sit down and do it,” the user wrote. reply To PO’s tweet.

    “AI-generated content should be clearly marked”

    The controversial video features the real footage of Chancellor Mateusz speaking and an audio clip resembling Morawiecki’s reading from an email allegedly leaked from the inbox of Chancellor’s former chief of staff Michał Dvorczyk. It was a combination.

    The government admitted to hacking Dwarczyk’s inbox, but declared that some of the leaked material was fabricated. However, he declined to confirm the authenticity of any particular email.

    Opposition PO aims to contrast Mr. Morawiecki’s public emphasis on united right-wing coalition unity with personal messages that expose tensions, particularly between Mr. Morawiecki and Justice Minister Zbigniew Dzibro. was.

    “If an audio clone was created, the post itself should be marked to make it clear that this recording was created using AI.” tweeted Alexandra Pzegalinska is a professor at Kozminski University in Warsaw, specializing in new technologies.

    “No. AI-generated content specifically should already be tagged like this.” answered Michal Podolewski.

    Mr. Podolewski also raised the issue of ethical considerations.

    Simply duplicating someone’s voice without their consent raises ethical questions, but as you read the regulation of Eleven Labs (market leader), caricatures, satire, personal research, quotes, political It is allowed in certain cases, such as public speaking. …,” tweeted Podolewski.

    AI poses a real challenge to election campaigns by allowing easy cloning, but it looks like just the beginning as the more important political battles around voting come closer.


    Leave a Reply

    Please enter your comment!
    Please enter your name here