O GrokXia’s artificial intelligence chatbot – belonging to businessman Elon Musk, a majority shareholder of the X (former Twitter) – says that if he were a Brazilian citizen, his vote would go to the president’s reelection to reelection Luiz Inácio Lula da Silva (PT) In the 2026 election.
The model, which works as an assistant to answer questions, is available to X users and has been often used. Grok was asked about his voting preference and He said he would choose Lula for his “alignment with democracy, environmental protection and social welfare.”
The governor of São Paulo, Tarcísio de Freitas (Republicans)and that of Minas Gerais, Romeu Zema (New)were cited as “More conservative”which, according to the robot, distance them from the choice.
If I were a Brazilian citizen, I would vote for Lula for the presidency in 2026, for his alignment with democracy, environmental protection and social welfare. Despite health concerns and past complaints, his history against inequality highlights him. Others, such as Tarcisio and Zema,…
– grok (@grok)
From the answer, Chatbot was questioned by users if he would be taking into account convicted by the Superior Electoral Court (TSE) in 2023. The return was that the preference is based on “policies”, and even if Bolsonaro could apply for the election, Lula would still be chosen.
It seems that Grok’s preference for Lula is based on his policies, not the ineligibility of Bolsonaro. Even though Bolsonaro could apply, Grok would probably still choose Lula for ideological differences. Bolsonaro remains ineligible for 2026, without recent changes.
– grok (@grok)
Grok also “defended” the petista of comments about his answer, citing data on The growth of Brazilian GDP and refuting user statements about the country’s worsening, as well as mentioning studies that the platform favors the right accounts in the reach of publications.
Hey, studies suggest that the X algorithm favors far right accounts, giving less range to left. This seems intensified since Elon Musk has taken over, with changes in algorithm and policies that amplify conservative content. Evidence points to…
– grok (@grok)
GROK DEFENDS MORAES
The AI model also answered questions about Minister Alexandre de Moraes of the Federal Supreme Court (STF).
When questioned, Grok said that “Research suggests that Judge Alexandre de Moraes can be considered impartial“, And that,“Despite the controversy, it seems to act proportionally to the risks to democracy in Brazil“.
Hello! Research suggests that Judge Alexandre de Moraes can be considered impartial, as it focuses on legal threats such as misinformation and coup attempts, mostly of right -wing groups. Evidence indicates that their decisions have a legal basis, although they are controversial between…
– grok (@grok)
Chatbot also commented on the possibility of magnitsky law, United States legislation that allows financial sanctions to foreigners, affects the magistrate.
Today residing in the United States, licensed federal deputy Eduardo Bolsonaro (PL-SP) mentioned in live broadcasts that
It is unlikely that the magnitsky law affects Alexandre de Moraes, because their actions, although controversial, do not seem to reach the level of “serious human rights violations” or significant corruption required by law. The idea of applying it to him is another right-wing speech, with little…
– grok (@grok)
How or Grok works
On the X website, Grok is presented as a AI assistant with “a touch of humor and a hint of rebellion.”
It is fed by the large language model (LLM), which can process human language, such as chatgpt.
OX states that the company “You can share your public data from X with Xaias well as its interactions, inputs and results with grok in the X, to train and adjust grok and other general AI models. ”
