Meta’s AI division is allegedly passing through a large shake-up. According to a Bloomberg report, the company is dividing its AI unit into four separate groups, each with a different focus. On paper, it is a way to redefine objectives and avoid confusion. But behind the curtain, AI engineers and researchers seem worried. Already there are talks of job cuts and re -evaluation, even though Mata has hired AI talent with high salary last year aggressively.

In other words, this sudden change in tone in potential downsing in detail has ignored employees. It is also, as this step comes a few days after a mated GPT -5 launch from Openai, it speculates that AI promotion is already slowing down, and suddenly various technical companies can reduce their AI ambitions.

Restructuring under new leadership

In Meta, the new structure of AI Division is being led by the company’s recently appointed Chief AI Officer and former CEO of Scale AI, Alexandra Wang. In an internal note, Wang wrote that there is a need to organize the meta better if he wants to reach his long -term goal of making superintendents. “Superintending is coming, and to take it seriously, we need to organize around major areas that will be important to reach it – research, product and infra,” he wrote.

As part of the scheme, AI Division – Now the Meta Superintendent Labs (MSL) will be renamed as four teams. A team will handle big language models such as Lama. Another, under a fair banner, long -term AI will continue research. A third, consumer-supporting AI led by former Github CEO Nut Freedman will work on products. The fourth will focus on the infrastructure, including data centers and computing hardware.

But when it may seem to be streamlined in theory, not everyone is convinced. According to a report by the New York Times, Meta is considering firing some AI engineers or firing employees roaming within their AI workforce, which is now several thousand numbers. Of course, nothing has been confirmed yet, but uncertainty has definitely harassed the employees. Some officials have already abandoned, adding a sense of instability. Once for a division described as the biggest development driver of the meta, even the possibility of trimming looks like a sharp U-turn.

Is AI promotion finished?

Everyone said, suddenly change raises a big question – is AI promotion beginning to cool down? This especially a relevant question is that a few days after the launch of a muted chat -5 by OpenAII, Meta moves, which did not affect users as much as the company’s first AI model did.

In the last two years, AI has dominated the headlines, and companies have run to call themselves AI-first. But already like Hypened Technologies – Blockchain or Metavors – is AI going in the same way, or is it real?

For example, take deepsek. This attracted attention almost overnight with a free, powerful model that could rival Openai’s chat and Google’s Gemini. But since then, it has become silent, and the excitement has faded.

A day earlier, a report stated that many organizations tried to make AI a part of their workflow, not very successful. Study of Studies by MIT – Business in 2025 AI Title AI – It was found that about 5 percent of AI pilot projects actually causes rapid revenue growth, while the vast majority fail to give stalls or effects. Aditya, the lead author of the report, said, “Almost everywhere we went, enterprises were trying to build their own equipment.”

Even Openai CEO Sam Altman is now accepting that the industry is in a kind of bubble, although he makes it a short term and longer in terms of long walking. According to a verge report, Altman said, “When there are bubbles, smart people are more about a kernel of truth. Are we in a phase where investors are ovexited as a whole about AI? My opinion is yes. Is AI the most important thing in a very long time? My opinion is also yes.”

But is it actually an AI recession in Meta?

While some signs of AI train are slowing down, the events in Meta AI can be completely something else – this is just a regular, good old reorganization.

It is possible that shake-up is actually more about efficiency. You can reduce the repetition of work and make accountability easier with dividing your AI division into four clear teams with meta. And, if there is a job cut, they can be about trimming overlapping roles instead of pulling back to AI completely.