The Meta article brings its AI to Trump course-not without the ulterior motive of Maria Gramsch first appeared on Basic Thinking. You always stay up to date with our newsletter.
Meta presented its latest AI language model with Llama 4. It should come more politically “balanced”. Critics fear a shift to the right, since meta boss Mark Zuckerberg is increasingly resisting the course of US President Donald Trump. That could too Meta AI affect WhatsApp, Instagram and Facebook.
At the beginning of April 2025, Meta presented its latest AI language model Llama 4. The artificial intelligence should According to the US group be more politically balanced.
In reality, this should be the same. Because Meta is based on Llama 4 on the AI chat bot Grok by Elon Musk. This is known for its conservative-populist tendencies, which often remind of the Musk’s attitudes. Meta boss Mark Zuckerberg has also been committed to the course of US President Donald Trump for some time.
What does Meta want to achieve with Llama 4?
Meta primarily wants to tackle bias with Llama 4. This bias of AI models goes back to the available training data and can influence results. According to Meta, AI models are “historically more oriented to the left”, especially “when it comes to debated political and social issues”.
This is also due to the training data available on the Internet. The US group therefore wants to “free his AI models from prejudices”. Llama should be able to “understand and articulate both sides of a controversial topic”.
Our tests show that Llama 4 answers a number of controversial political or social topics with a GRK -comparable rate with strong political tendency.
Llama 4 can now better go into “split questions” – without taking a party. The AI model can react to questions with “a variety of different points of view” without judging.
On Trump course?
Meta emphasizes the desired “balance” in its publication, but experts see this critically. “It is dangerous to consider scientific and empirical questions such as climate change, health or environment with political glasses as a left/right-oriented manner,” said Ababa Birhan from the Mozilla Foundation opposite 404 Media.
The “both sides” approach represents a false equivalence. Because that is as if you were equivalent to a vaccination opponent theorist with a scientist or doctor. “One is illegitimate and dangerous, the other is based on verifiable empirical evidence,” said the AI expert at Mozilla Foundation.
The procedure shows that Mark Zuckerberg wants to move his company-and also his AI model. The background: Mark Zuckerberg tries to orient Meta to Donald Trump’s policy, since the US President apparently includes the big tech corporations.
This could be accompanied by META because the company is currently on trial. The US cartel authority FTC accuses the US group of establishing a monopoly in the social media sector and suppressed the competition by taking over the takeover of Instagram (2012) and WhatsApp (2014). The authority therefore requires Meta to separate from both platforms.
Previously, the company had already completed a political U -turn and adapted its community rules for the United States. On Facebook, Instagram and threads, he allegedly wants to work for freedom of speaking – and therefore check fewer facts. Mark Zuckerberg is obviously hoping for an advantage in the trial of meta in court.
What influence does Llama 4 have on the AI assistant Meta AI?
For a few months now, Meta AI has been available on the meta platforms of the AI assistants. On Facebook, Instagram, WhatsApp, Threads and on Facebook Messenger, users can now chat with the AI bot. But what do the changes at Llama mean for the Meta AI? Can a right to be expected on Facebook, Instagram and WhatsApp?
So far, Meta AI in the EU is currently still running with Llama 3.2. The new Llama 4 model is already available in 40 countries, but not yet in the European Union. The US group had so far set up the publication due to regulatory uncertainties.
However, if Llama 4 will also be published in the EU in the future and are used in the AI assistant Meta AI, the new “balance” of the AI model could be clearly noticeable in the answers.
Also interesting:
- Meta: AI training with EU user data-this is how you can object
- About the company construct: Tesla apparently prevents taxes in Germany
- Overview with AI: Google Ki lets traffic of German websites break in
- Ex-employees’ allegations: Meta is said to have spy for China
The Meta article brings its AI to Trump course-not without the ulterior motive of Maria Gramsch first appeared on Basic Thinking. Follow us too Google News and Flipboard.
As a Tech Industry expert, I would approach this situation with caution and skepticism. While Meta may claim to be bringing AI to a course on Trump, it is important to consider the potential ulterior motives behind this decision.
Meta is a company that has faced significant scrutiny and criticism for its handling of user data and privacy concerns. Bringing AI to a course on Trump could potentially be a way for Meta to gather more data on user behavior and preferences, furthering their own agenda and potentially exploiting users for their own gain.
Additionally, the use of AI in a course on Trump raises concerns about bias and manipulation. AI algorithms are not immune to bias and can easily be manipulated to serve a particular agenda. It is crucial to consider the potential impact of using AI in this context and ensure that it is being used in an ethical and responsible manner.
Overall, while the idea of bringing AI to a course on Trump may seem innovative and cutting-edge, it is important to approach this development with a critical eye and not overlook any potential ulterior motives that Meta may have.
Credits