'Don't Be Evil'? Whistleblower Claims Google AI Assisted Israel's Military Amid Gaza Conflict

A former Google employee has raised concerns about the tech giant's involvement in military operations. The whistleblower claims that artificial intelligence is being used to monitor and identify individuals in conflict zones. These allegations have sparked a fresh debate regarding the ethical boundaries of Silicon Valley's partnerships with defence forces.

A former Google staffer has filed a confidential federal whistleblower complaint, seen by The Washington Post, alleging the tech firm flouted its 2024 ban on AI-assisted warfare and tracking. The report claims the company assisted an Israeli military contractor with the analysis of drone recordings, directly contradicting its own ethical policies.

Internal files from the complaint indicate that Israel's military was using Google's Gemini AI at the very moment the company was publicly downplaying its ties to the country's armed forces. This surfaced after staff staged protests regarding a specific contract with the Israeli government.

Identifying the Contractor

Files from an August SEC complaint reveal that a user with an Israel Defence Forces (IDF) email address contacted Google's cloud division for help in July 2024. This customer's name appears to match that of an employee at an Israeli firm called CloudEx, which the whistleblower identifies as a third-party provider to the military.

Internal records from the filing show that a user from the IDF email account sought advice on improving Gemini AI's ability to detect drones, tanks, and troops in aerial footage. In response, Google's cloud team provided technical guidance and conducted internal trials to refine the technology.

Google's public ethical guidelines at the time specifically prohibited the use of AI for weapons or for monitoring that violated international rules. According to the complaint, the support provided to the Israeli military contractor ran counter to both core policies.

The SEC filing alleges that Google violated securities regulations by ignoring its own public rules. Because these policies were also part of official government reports, the whistleblower argues that the company effectively lied to its shareholders and the authorities.

In a statement to The Post, the former staffer—who requested anonymity to avoid company reprisals—explained that while Google's internal ethics reviews are usually 'robust', those standards seemed to vanish regarding the conflict in Gaza.

'Many of my projects at Google have gone through their internal AI ethics review process,' they noted, adding that they approached the SEC to ensure the firm is 'held accountable for this double standard.'

Google spokesperson challenged the claims, stating that the company's AI principles remained intact. They argued that the account's activity was too minor to be 'meaningful', noting that the support request came from a user spending less than a few hundred dollars a month—an amount they claimed made any significant use of AI impossible.

The 'Double Standard'

In a statement, a Google spokesperson disputed the claims, asserting that the company provided only 'standard, help desk information' for a routine query and offered no specialised technical support. They further argued that because the account spent less than a few hundred dollars a month on AI, any 'meaningful usage' of the technology for military purposes would have been 'impossible.'

According to Google's own guides for its 'cloud video intelligence' tools, the service allows users to track objects in footage at no cost for the first 1,000 minutes. After the initial period, the company charges 15 cents per additional minute of video processed.

Public records for Google's 'cloud video intelligence' tools show that object tracking is offered at no cost for the initial 1,000 minutes. Once that threshold is crossed, the firm charges 15 cents per additional minute, a rate that can support significant analysis even on a modest budget.

A spokesperson for the SEC declined to comment on the matter. While the agency allows anyone to lodge a grievance, these filings are kept private and do not automatically trigger an official probe. Representatives for CloudEx and the IDF did not provide a statement.

The 'IT for IDF' Connection

In 2024, CloudEx was listed as a sponsor for 'IT for IDF', a technology summit held south of Tel Aviv. According to the event's website, the conference featured Israeli military leaders who spoke on the vital role cloud computing plays in their current operations within Gaza.

While the whistleblower's complaint alleges that the use of Gemini AI was connected to operations in Gaza, it provides no concrete evidence for the claim. This stands in contrast to Google's prior statements, in which the firm insisted that its Israeli contracts avoid 'military workloads relevant to weapons or intelligence services.'

Since its inception, Google has focused on serving the general public, a mission once famously summed up by its 'Don't be evil' slogan. More recently, the firm's pursuit of military partnerships has created a rift within the company, as employees have taken a stand against its shift toward defence work.

The company established its AI guidelines, which prohibit weapons and surveillance work, in 2018. This move followed a major internal revolt, during which Google withdrew from a Pentagon contract that used machine learning to process drone footage.

In 2018, after intense pressure from staff, Google cancelled a project for the US military involving drone clip analysis and introduced its current AI principles. These rules were designed to prevent its technology from being used for tracking or the development of armaments.

Industry-Wide Tensions

In 2021, Google and Amazon won a $1.2 billion (£0.88 billion) cloud deal with Israel's government known as Project Nimbus. Microsoft also provides cloud computing to the Israeli government.

Internal opposition to Israeli government contracts has grown across Google, Amazon, and Microsoft. According to a workers' group, Google fired more than 50 employees in April 2024 after they occupied company offices to protest the collaboration. Microsoft has taken similar action, dismissing employees who participated in demonstrations at its campus outside Seattle.