Google DeepMind staff are against military contracts for AI use

Hotstar in UAE
Hotstar in UAE

In its long-term bet on the AI segment, Google bought DeepMind in 2014. Today, the latter is deeply integrated into the company’s organizational chart, losing autonomy. However, there is some internal dissent regarding the handling of some issues. Mainly, a section of Google employees at DeepMind are against military contracts.

The concern stems from Project Nimbus, a $1.2 billion contract where Google would provide AI-powered services and tools derived from DeepMind’s work to the Israeli military. Google is not alone in this, as Amazon is also involved. In mid-May, 200 DeepMind employees signed a letter saying they are “concerned by recent reports of Google’s contracts with military organizations.”

200 Google DeepMind employees signed a letter against military contracts

While Project Nimbus was first heard of in 2023, employee protests against it rose this year. Reports from April claimed that Google fired 28 employees who participated in the protests. Tensions seem to have been escalating since then, although there were developments that were kept secret. For example, a recent TIME report revealed the existence of the letter mentioned above. Although 3 months have passed, people involved with the matter say that they have “received no meaningful response from leadership.”

The main concern of DeepMind employees is that their technology would be used for military purposes. “Any involvement with military and weapon manufacturing impacts our position as leaders in ethical and responsible AI, and goes against our mission statement and stated AI Principles,” reads the letter. Employees feel that Google is “unspecific” or vague when referring to Nimbus. They even feel that there are clauses that would allow the company to supply its technology for military purposes.

DeepMind lost a lot of independence

Currently, DeepMind and Google Brain are one AI-focused team. The merger occurred in 2023 after a DeepMind attempt to seek more independent management, something they had at the beginning. However, DeepMind now follows a general Google ethics policy. Part of that policy says the company can develop technologies to cause harm if “the benefits substantially outweigh the risks.” But it is not clear who or how Google determines whether or not to cross the line.

Meanwhile, a Google spokesperson said Nimbus is not what people think. An official statement said the project “is not directed at highly sensitive, classified, or military workloads relevant to weapons or intelligence services.” Meanwhile, DeepMind employees feel they can’t do much about it. “While DeepMind may have been unhappy to work on military AI or defense contracts in the past, I do think this isn’t really our decision any more,” one of them told TIME.

Staff doubtful about how DeepMind technology is actually being used

Currently, it is only certain that Israel and its Ministry of Defense have access to Google’s Cloud software. The Mountain View giant has integrated AI-powered technologies into the latter, as in many of its services. The United States would also be forcing Israel to buy artificial intelligence technology from Google and Amazon. However, Google’s position is that the software is not being used for military purposes or weapons manufacturing. Sadly, there is no way for company employees to corroborate if it is true.

2024-08-23 15:08:53