The End of Technological Autonomy: Meta Relies on Midjourney, Apple Looks to Gemini to Relaunch Siri

While Meta is buying "aesthetics," Apple is banking on "intelligence. " Siri, launched in 2011 as the first global voice assistant, has stagnated . While Alexa and Google Assistant were gaining ground, the arrival of generative chatbots like ChatGPT has highlighted the backwardness of a system incapable of truly understanding context or sustaining natural conversations.

An attempt to reverse this trend came with Apple Intelligence , announced in 2024, with models that operate directly on devices and, in the most complex cases, on private cloud servers to ensure data confidentiality. However, the limitations were immediately clear: Siri is not yet able to compete with its rivals.

This is why Apple is discussing with Google the use of Gemini , Mountain View's flagship model. According to rumors collected by Bloomberg and Reuters , the idea is to develop a customized version of Gemini, hosted on Apple infrastructure, so as to combine conversational power and privacy protection . The new Siri could arrive starting in 2026, according to Bloomberg , but the decision must be made soon. In the meantime, Samsung has already brought Galaxy AI to its devices, combining its Samsung Gauss model with solutions like Gemini, while Microsoft has transformed Copilot into an assistant across Windows and the Office suite. The risk for Cupertino is that Siri, even enhanced, will arrive late in an already crowded market .

Meanwhile, however, the hypothesis has already heated up the markets : Alphabet shares have risen by more than three percentage points, Apple's by about one and a half, according to Reuters . If the deal goes through, it would be a momentous change for the giant led by Tim Cook, which has always made end-to-end control its hallmark.

It's not just Meta and Apple. In recent times, the geography of AI has been reshaped by a logic of alliances involving all the major players : Microsoft has invested billions in OpenAI, Amazon has formed a strategic partnership with Anthropic. In each of these cases, the logic is the same: computing costs have exploded, datasets require unprecedented resources and authorizations, and innovation cycles have shortened to the point that not even the richest companies can afford to remain locked in their labs.