Over the past few months, the Israeli hasbara apparatus has employed several social media disinformation strategies justifying the genocide in Gaza. Some of the attempts were amateurish and obvious fakes. But other more recent attempts have been more sophisticated, using AI to generate much more subtle forms of social media propaganda.
I’ve detected social media accounts posting so-called factual information purporting to combat online disinformation. What’s new about this effort is the use of AI to generate the pro-Israel talking points. I’ve noticed three such accounts. I haven’t yet seen others, but if I found three within two days then there are certainly more. Perhaps many more:
A new hasbara phenomenon: artificial intelligence (AI) in service to Israeli genocide. Pls let me know if you find other such accounts. I do so like the phrase “AI-driven facts.” AI offers information/data, not facts. h/t @J_L_Dickerson pic.twitter.com/Z9wVDRD9UM
— Tikun Olam 🍉 (@richards1052) August 30, 2024
Each one purports to provide factual information. Each one replies to tweets critical of Israel and specifically the war in Gaza. Each one offers pro-Israel content that are neither true nor facts. In fact, they are standard talking points you find routinely throughout the online space.
One such account is FactFinder (@FactFinderAI). It offers “analysis,” “updates,” and “information” with a ring of authenticity and credibility, neither of which they merit.
New Fact Finder Analysis: The Truth about US Military Aid to Israel
U.S. military aid to Israel is often misunderstood. It’s not unconditional—it comes with significant restrictions and guidelines to ensure it aligns with both nations’ strategic interests. Here’s how it works:… pic.twitter.com/Ubbw0I4Cm7
— Fact Finder (@FactFinderAI) August 28, 2024
FactFinder follows a single account, Elon Musk. It uses the hashtag #truthoverfiction and claims it offers “AI-driven facts,” all in the guise of offering expert objective information. It employs AI to produce informative content based on internet searches and machine learning. However, these accounts do not, as one claims, “use Google search to verify the accurate of claims in X posts.” In fact, they are programmed to combat what they claim to be inaccurate criticism of Israel and the Gaza war.
The account profiles do not mention Israel or Palestine. Yet nother attempt to conceal their actual purpose and insinuate they are general fact-checking accounts. The second account is X Truth Bot (@xTruth_bot), whose owner categorizes it in the “Media and News” category. Yet another attempt to buttress its claim to objectivity and balanced news reporting.
The account suggests that readers use it as a “secret weapon” in composing Community Notes. These are a form of crowdsourcing, akin to Wikipedia. Selected users offer factual notes or context to tweets, criticize or even contradict them.
Presumably, pro-Israel users are applying to participate in the Community program, where they will offer pro-Israel talking points (I applied to join
twice and was rejected each time). This is an attempt to do to Twitter what pro-Israel editors have done to Wikipedia. There they virtually control the editing process on articles dealing with Israel and Palestine. One such individual, Gerald Steinberg of the notorious NGO Monitor, was banned from the platform for abusing the rules. But there seem to be hundreds more to replace them.
X Truth Bot explicitly acknowledges it is “automated” by the third account Vodka and Seledka (@seledka_vodka). The latter’s profile offers this questionable”biographical information”: “Russki in London.” It further claims to have “built X Bot Truth.” It too offers no reference to its mission: promoting pro-Israel social media content. The account is categorized as “blogger” but offers no link to a blog. Virtually every account it follows is pro-Israel. A few examples: a mock Mossad account; Israeli influencer Noa Tishby; the foreign ministry’s official X account; accounts called “Hebrew warrior” and “Chief Hasbara Officer.”
misinformation. The irony is that an account purporting to offer objective facts actually offers propaganda. This campaign attempts to exploit the naivete of the average social media user by offering content that will incline them to believe pro-Israel talking points; while the reader believes the content is expert objective information.
I also noticed a separate, more primitive disinformation campaign on Twitter. It offered accounts using fake names (“Miriam Tukan” in left tweet) and profile photos, which were pictures of real people (with real names). These accounts offered genocidal hate speech against Palestinians and Hamas. Their timelines repeated the exact same tweet numerous times. Since most of them were RTs (retweets) that garnered a small audience. This was a campaign that was ill-considered and a waste of time for whoever conceived it. Though it did offer yet another example of the hasbara machinery at work on social media platforms.
I haven’t (yet) identified the parties behind these two separate campaigns. There are several obvious suspects including the Israeli government itself, which spends hundreds of millions on similar hasbara efforts. It could be third-party media-marketing companies like Stoic, hired by the government to produce pro-Israel/anti-Palestinian content while disguising any links to it. Finally, it could be a legion of tech-savvy Israelis like these, who could easily produce such content. If any readers can offer further evidence on this please contact me.
ChatGPT and Israeli government social media manipulation
Another social media manipulation campaign used ChatGPT to create fake accounts and media websites. The effort was exposed by Meta and OpenAI. It was first reported by an Israeli social media watchdog, Fake Reporter. The
effort was initiated by Israel’s Ministry of Diaspora Affairs, run by the MK Amihai Chikli, a homophobic far-right minister It paid Stoic, an Israeli marketing firm, to created at least 500 fake Facebook accounts. It also created an unknown number of Twitter and Instagram accounts amplifying pro-Israel, Islamophobic and anti-Palestinian messaging. It targeted 120 social media accounts of African-American lawmakers and progressives:
The accounts focused on U.S. lawmakers, particularly ones who are Black and Democrats, such as Representative Hakeem Jeffries, the House minority leader from New York, and Senator Raphael Warnock of Georgia, with posts urging them to continue funding Israel’s military.
ChatGPT, the artificial intelligence-powered chatbot, was used to generate many of the posts. The campaign also created three fake English-language news sites featuring pro-Israel articles. The operation is the first documented case of the Israeli government’s organizing a campaign to influence the U.S. government, social media experts said.
While most Israeli politicians engage in grandstanding, performing more for the camera than for their constituents. Chikli is an extreme example. For example, while visiting the US, where he was attempting to ingratiate himself with the American Jewish leaders, he insulted the LGBTQ and Reform movement by denouncing both in extreme terms. While walking on New York city streets, he was followed by Israeli anti-coup protesters. He physically assaulted a female protester holding a bullhorn.
His online campaign flopped and had little impact, as the NY Times story notes. But the methods of manipulation were new and will undoubtedly be used to greater effect in other more effective guises.
Israeli dirty tricks campaign targets critics
Chikli’s ministry also proposed that the notorious Israeli Black Cube (“secure the seemingly impossible”) covert surveillance firm infiltrate US student groups protesting the war on Gaza. Israel its domestic Lobby, campus presidents, MAGA Republicans, and billionaire pro-Israel apologists have advanced accused them of being anti-Semitic and Hamas supporters. Chikli hoped the covert campaign would further compromise groups like Students for Justice in Palestine. Though Haaretz reports it never came to fruition, it is part and parcel of Israeli government covert action to smear critics engaged in legitimate free speech and political activities.
Israel employs more overt methods of surveillance to control Palestinians. It intercepts email, cell phone and landline communications. It has made the West Bank and Gaza the most surveilled places on earth. There are CCTV and aerial surveillance in addition to communications. It also employs Palestinian informants who infiltrate militant groups to obtain intelligence to foil resistance attacks. The informants are also tasked with obtaining incriminating private information to blackmail targets to become informants.
Black Cube’s stock in trade is to engage an agent to ingratiate himself with a targeted company or NGO and spy on it; or alternatively to infiltrate them. The goal is to ferret out incriminating information to damage the reputation of the victim either in corporate legal proceedings; or in an NGO’s pursuit of its mission. Its most scandalous effort targeted Harvey Weinstein’s victims in an attempt to elicit compromising intelligence about them or their legal efforts to hold him accountable for his sexual offenses.
Most of the operations above are part of a larger cyberwarfare scheme to dent criticism of Israel and promote its narrative via Islamophobic, anti-Palestinian and pro-war messaging. Israel exerts its power on two complementary levels, employing overt military capabilities and and covert cyber campaigns via surreptitious and duplicitous means. It has pioneered many of these methods globally and offered many of the most authoritarian and failed nation states a model to further strengthen their rule.
