Deepfakes, bots and disinformation: AI in the Israeli election campaign
The Israeli presidential election campaign has been significantly influenced by artificial intelligence, with researchers at the Open University of Israel documenting over 2,000 AI-generated posts from Prime Minister Benjamin Netanyahu's Likud party alone. A notable example involves a deepfake video posted by Netanyahu's account showing candidate Gadi Eisenkot embracing an Arab politician rather than his deceased son, which garnered over one million views and sparked public outrage. The ZebeAI research project analyzed more than 52,000 social media posts across various platforms, finding that over 14 percent utilized AI technology, often through hybrid methods involving images, music, or subtitles. Political scientists describe this as the first Israeli election heavily shaped by AI, where parties employ automated calls, bots, and provocative videos to sway voters. Experts warn that such technologies could be strategically deployed close to election day to stir controversy and potentially impact final results.
On Thursday, Israeli authorities disputed Prime Minister Benjamin Netanyahu’s assertions that the recent legislative election was compromised by widespread fraud in Arab communities. While the head of the Central Election Committee requested a security investigation into allegations of foreign interference, police and election officials stated there is no evidence to support claims that the vote was stolen. The dispute erupted as final results indicated that Netanyahu’s coalition had lost its parliamentary majority, leaving both the prime minister and his main opponent, Gadi Eisenkot, short of the seats needed to form a government without cooperation from Arab-led parties. The crisis began following Tuesday’s election, where Arab parties Ra’am and the Joint List secured a record nineteen seats, driven by unprecedented voter turnout. Likud, Netanyahu’s party, initially cited exit polls suggesting a setback for the prime minister to launch a campaign questioning the validity of the vote. By Wednesday, Likud issued a statement declaring the elections “stolen by Arab parties that support terrorism and the elimination of the Jewish state.” The party alleged that severe criminal offenses, including bribery, threats, violence, and forgery, were used to mobilize voters. Likud claimed to possess hundreds of testimonies and thousands of envelopes containing cash, though it did not publicly release this evidence. The situation intensified on Thursday when Noam Sohlberg, the Supreme Court judge presiding over the Central Election Committee, sent a letter to the Shin Bet, Israel’s internal security agency. Sohlberg wrote that serious suspicions of improper intervention by foreign actors acting against Israeli interests had been brought to his attention. Although he stated he was unable to verify the truth of the allegations, he instructed the Shin Bet to investigate thoroughly and submit conclusions within a week. The letter was subsequently leaked to Israeli media, identifying Qatar as the specific foreign entity alleged to have funneled money to an Arab party to boost turnout. In response to the leak, Likud labeled the incident the “Ra’amgate Affair,” drawing a parallel to historical political scandals. The party argued that the alleged interference aimed to facilitate a foreign takeover of the Knesset, the Islamization of Israel, and the overthrow of right-wing rule. National Security Minister Itamar Ben-Gvir called for the arrest of leaders of Arab parties suspected of receiving Qatari funds, as well as figures linked to the Muslim Brotherhood. However, Israeli police contradicted these claims, stating there was no investigative or informative evidence to support the notion of a stolen election. Police noted that election day proceeded with only a minority of irregular incidents, comparable to previous cycles. Dean Livne, the acting director general of the Central Election Committee, echoed the police assessment in an interview with public broadcaster Kan 11. He stated that the committee had received claims rather than evidence from Likud and other parties, asserting there was no basis for accusations of a stolen election. Meanwhile, Qatar issued a statement denying any involvement in the Israeli electoral process, describing the allegations as false propaganda intended to drag the nation’s name into Israel’s internal political conflicts. Gadi Eisenkot, the head of the Yashar party and Netanyahu’s primary competitor, condemned the fraud allegations as an improper influence operation designed to create a fictitious delegitimization of the upcoming parliament. Eisenkot stated that Likud was attempting to cast doubt on the results because they had lost, while simultaneously calling for the specific allegations to be investigated. He warned that Netanyahu was positioning himself to block any potential alliance between the opposition bloc and the Arab parties, suggesting such cooperation would be treasonous. The controversy unfolds against a backdrop of intense polarization and the integration of artificial intelligence into the campaign. Researchers from the Open University of Israel’s ZebeAI project tracked over 52,000 social media posts by politicians between late July and October 23, finding that more than 14 percent contained AI-generated content. Likud accounted for the highest absolute number of AI-assisted posts, using deepfakes and automated bots to stir controversy. One notable example was a 19-second video shared by Netanyahu’s account, depicting Eisenkot ignoring his deceased son to embrace Mansour Abbas, the chairman of Ra’am. The clip, which played over one million times, sparked outrage given that Eisenkot’s son died in the 2023 war. Despite new regulations requiring labels for AI-generated content, experts noted that legal frameworks remain outdated, allowing practices that would be restricted in Europe. Political analysts suggest Likud’s strategy aims to reduce Arab influence in Israeli politics and deter the opposition from forming a governing coalition with Arab parties. Amira Halperin, a senior fellow at the Palm Beach Center for Democracy and Politics, noted that decreasing Arab influence appears to be a key motivation. The projected outcome shows both the religious-nationalist bloc led by Netanyahu and the secular-liberal opposition falling short of a majority. Consequently, either side would need to negotiate with the Arab parties to secure power. As final results confirmed the loss of the majority for Netanyahu’s camp, the debate shifted from the validity of the count to the feasibility of future coalitions, with the Shin Bet investigation set to conclude within the mandated timeframe.
Go to the primary sources (3)
The official sources this coverage is built on. Read them directly to bypass framing.
Following Israel's recent legislative election, Prime Minister Benjamin Netanyahu's Likud party alleged that foreign actors influenced voter mobilization in Arab communities through bribery and threats. The party claimed the elections were 'stolen' by Arab parties, specifically Ra'am and the Joint List, which achieved record-high turnout among citizens of Arab descent. In response, No'am Sohlberg, the Supreme Court judge heading the Central Election Committee, requested the Shin Bet internal security agency to investigate these specific claims. However, Israeli police stated there is no investigative or informative evidence supporting the allegations of large-scale fraud. A senior official from the Central Election Committee also refuted the accusations, noting that while some minor irregularities occurred, there is no basis for the claim that the election results were compromised.
Bias read (Center): The article presents a balanced view by detailing the specific allegations made by the Likud party and immediately contrasting them with the direct denials from the police and the Central Election Committee. It includes quotes from both the accusers and the authorities responsible for the election,
Why factuality (85): The article accurately reports the specific political event regarding the 2025 Israeli elections, citing the Supreme Court judge's letter and Likud's statements. It faithfully represents the primary source's definition of propaganda by highlighting how one side uses selective information (blinding o
Why objectivity (90): The tone remains neutral and journalistic. It presents the accusations of fraud and the official response without adopting a partisan stance. It attributes views clearly to the respective parties (Likud vs. the Election Committee) and avoids loaded language.
The Israeli presidential election campaign has been significantly influenced by artificial intelligence, with researchers at the Open University of Israel documenting over 2,000 AI-generated posts from Prime Minister Benjamin Netanyahu's Likud party alone. A notable example involves a deepfake video posted by Netanyahu's account showing candidate Gadi Eisenkot embracing an Arab politician rather than his deceased son, which garnered over one million views and sparked public outrage. The ZebeAI research project analyzed more than 52,000 social media posts across various platforms, finding that over 14 percent utilized AI technology, often through hybrid methods involving images, music, or subtitles. Political scientists describe this as the first Israeli election heavily shaped by AI, where parties employ automated calls, bots, and provocative videos to sway voters. Experts warn that such technologies could be strategically deployed close to election day to stir controversy and potentially impact final results.
Bias read (Center): The article maintains a neutral, investigative tone by relying on data from academic research projects like ZebeAI and quoting experts on the technical usage of AI in campaigning. It presents specific examples of controversial content but does not take a stance on the political merits of the parties
Why factuality (80): The article describes a specific incident involving AI-generated content in the 2025 election campaign. While the primary source document defines propaganda generally rather than detailing this specific video, the article's description of using media to manipulate emotions aligns perfectly with the
Why objectivity (85): The writing style is descriptive and analytical. It explains the mechanism of the propaganda (using AI to evoke anti-Arab sentiment) without explicitly condemning the actor, though the framing naturally highlights the manipulative aspect. It maintains a professional distance from the political confl
This article examines the evolution of state propaganda in the era of artificial intelligence, using the hypothetical 2026 US-Iran conflict as a case study. It highlights how AI-generated content, such as animations depicting Iranian military power or US Vice President JD Vance destroying Tehran, serves as modern propaganda tools that have garnered millions of views. The piece defines propaganda as the selective presentation of information to influence behavior, noting that while authoritarian states centralize these efforts, democratic nations utilize more pluralistic but still manipulative methods. Expert commentary from political scientist Piers Robinson illustrates this through the 2003 Iraq War, where US politicians allegedly used disinformation regarding weapons of mass destruction to justify military action. The text concludes by discussing how propaganda aims to discredit political opponents and how figures like Donald Trump have utilized AI imagery to shape their public image.
Bias read (Center): The article adopts an educational and analytical tone typical of fact-checking initiatives. It explicitly contrasts the centralized nature of propaganda in authoritarian regimes with the pluralistic nature of propaganda in democracies, avoiding a one-sided critique of any specific political system.而
Why factuality (40): This article contains significant factual errors relative to reality and likely the intended context. It references a "US-Iran War in 2026" and mentions JD Vance as Vice President in a war scenario, which appears to be fictional or speculative future fiction presented as current news fact-checking.
Why objectivity (95): Despite the factual issues, the tone is highly objective and educational. It adopts a neutral, explanatory voice to define propaganda and distinguish between authoritarian and democratic contexts, mirroring the primary source's definitions closely without injecting personal bias or emotional languag
How each side covered it
The same event, grouped by the political lean of the outlets covering it.
progressive
center
conservative
★
How each side covered it
Support independent, bias-aware news and unlock the social pulse, community voting, and every other Supporter feature.
Germany
