A protester rose in defiance at a conference promoting the Israeli technology sector where Google’s managing director for Israel, Barak Regev, was delivering a speech in downtown Manhattan on March 4. The activist, identified as a Google Cloud program expert, vehemently expressed refusal to participate in the development of technologies that support genocide, racism, or surveillance. Clad in an orange t-shirt adorned with a subtle Google logo, the protester declared, “There is no place for technology that fosters segregation!”
In a recorded video of the incident, the Google employee, a 23-year-old software engineer named Eddie Hatfield, faced jeers from the crowd and was swiftly escorted out of the vicinity. Following a brief pause, Regev acknowledged the act of dissent, stating, “Embracing diverse viewpoints is a hallmark of being part of an organization that champions democratic values,” addressing the gathered audience.
Subsequently, Hatfield was terminated from his position at Google three days later.
Hatfield is part of a burgeoning movement within Google advocating for the termination of Project Nimbus, a $1.2 billion collaboration with Israel shared concurrently with Amazon. The opposition group, No Tech for Apartheid, reportedly comprises approximately 40 Facebook employees actively engaged in advocacy, with claims of hundreds more employees aligning with their mission. TIME magazine interviewed five current and former Google employees for this account, many of whom expressed mounting apprehension regarding Google’s potential involvement in Israel’s conflict in Gaza. Notably, two original Google staff members disclosed their resignations from the company in protest against Project Nimbus. The details surrounding Hatfield’s identity and these resignations have not been publicly disclosed.
The focus of No Tech for Apartheid’s campaign centers more on undisclosed aspects of Project Nimbus rather than its explicit objectives. As per the Israeli finance ministry’s announcement in 2021, the agreement entails Google and Amazon providing AI and cloud computing services to the Israeli government and military. The utilization of Google’s technologies ostensibly extends to empowering the Israeli government in conducting extensive data analysis, Artificial Intelligence training, database hosting, and other computational tasks on Google’s infrastructure. The Intercept initially reported in 2022 on Google’s purported involvement in facilitating Israel’s Cloud services, including capabilities such as AI-enabled facial recognition, image classification, and object tracking.
The specifics of the contract remain opaque, with employees voicing concerns over Google’s alleged lack of transparency regarding the full scope of Project Nimbus and the nature of its association with Israel. The exact functionalities offered to Israel under the contract have not been disclosed by Israel, Amazon, or Google. A Google spokesperson clarified that the Nimbus contract primarily caters to workloads managed by Israeli government ministries, excluding sensitive or classified military operations related to weaponry or intelligence services. The spokesperson emphasized Google Cloud’s adherence to terms of service and acceptable usage policies, prohibiting any activities that contravene laws or incite violence resulting in harm to individuals. Amazon echoed a similar sentiment, emphasizing the company’s commitment to providing cloud services universally while supporting employees affected by conflicts and collaborating with humanitarian organizations. At the time of reporting, the Israeli government had not issued a response to requests for comments.
Notably, there is no concrete evidence linking Google or Amazon’s technologies to civilian casualties. The dissent among Google employees stems from three primary concerns: the explicit acknowledgment by the Israeli finance ministry that Nimbus would serve the ministry of defense, the potential applications of Google’s cloud services within the Israeli government, and Google’s purported inability to oversee Israel’s utilization of their technology. Employees fear the misuse of Google’s advanced AI and cloud computing capabilities for surveillance, military targeting, or other offensive purposes. The terms of the contract reportedly restrict Google and Amazon from intervening in the usage of their services by specific government branches, including the Israeli military, irrespective of public pressure.
Recent reports in the Israeli media suggest that AI-driven targeting systems are being employed in airstrikes, raising questions about the cloud provider facilitating such operations. According to Google employees, tech companies typically maintain limited oversight over the activities on sovereign cloud servers of government clients for security reasons. Concerns linger regarding the potential misuse of technology for military endeavors, prompting employees to question the accountability mechanisms in place to prevent such violations.
No Tech for Apartheid members assert that their movement is gaining momentum in response to revelations concerning AI’s involvement in Israel’s military operations in Gaza, recent incidents of Israeli military actions against foreign aid workers, and global calls for a ceasefire. The precedent set by Google employees successfully influencing the company to retract a separate Pentagon contract in 2018 underscores the growing discontent within the workforce. Against the backdrop of escalating international criticism of the collateral damage inflicted by Israel’s actions in Gaza, many employees perceive Google’s dismissal of Hatfield as an effort to quell dissent and safeguard its business interests. Hatfield, speaking publicly for the first time, believes his termination was intended to stifle the burgeoning resistance within Google. He remarked, “I interpret Google’s decision to fire me as a signal of their concern over the traction this movement is gaining. It was a calculated move to set an example and deter further dissent.”
Hatfield disclosed that his protest was the culmination of internal inquiries directed at Google leadership regarding Project Nimbus, where he felt his concerns were disregarded. He recounted a meeting with his manager and an HR representative three days post the conference disruption, during which he was informed that his actions had tarnished the company’s reputation, leading to his immediate dismissal. In response, a Google spokesperson asserted that Hatfield’s disruptive conduct at an official company event violated company policies, justifying his termination as a consequence of unacceptable behavior.
Vidana Abdel Khalek, witnessing Hatfield’s dismissal, decided to tender her resignation in protest against Project Nimbus on March 25. In an email addressed to company executives, including CEO Sundar Pichai, Khalek condemned Google’s involvement in offensive military technology, citing the devastating impact on civilian lives amid ongoing conflicts. Khalek’s departure underscores the ethical dilemmas faced by employees grappling with the discrepancy between Google’s stated principles and its actual engagements. The discord between Google’s AI principles and its collaboration with Israel prompts Khalek to question the company’s neutrality and complicity in enabling harmful practices.
The restructuring within Google DeepMind, the company’s AI division, has raised concerns among employees regarding the lab’s autonomy in preventing the militarization of its AI tools. While DeepMind initially established safeguards against military and surveillance applications, subsequent governance changes aligned its operations with Google’s overarching AI principles. This alignment has implications for Google’s pursuit of lucrative contracts with entities like Israel and the Pentagon, potentially compromising the ethical standards initially upheld by DeepMind. Employees express apprehensions over the lack of clarity regarding the application of these principles to projects developed by DeepMind and subsequently deployed via Google Cloud to clients, including governments like Israel.
Employees familiar with Google Cloud technology highlight the potential risks associated with Project Nimbus, particularly concerning the AutoML tool that enables rapid machine learning model training. The Israeli government’s hypothetical utilization of AutoML for surveillance or targeting purposes underscores the ethical quandaries faced by Google employees. While the specifics of Israel’s use of facial recognition technology remain unconfirmed, concerns persist over the ethical implications of providing advanced technological capabilities to entities with a track record of weaponizing AI for military objectives.
Members of No Tech for Apartheid stress the imperative of acknowledging the potential misuse of Google’s technology by the Israeli government for coercive ends. The prevailing apprehensions revolve around the absence of stringent oversight mechanisms to prevent the weaponization of AI and cloud computing resources provided by Google. The reluctance of employees to openly criticize Project Nimbus or voice support for Palestinians stems from fears of reprisal or professional repercussions, underscoring a culture of silence and apprehension within the organization.
Despite the challenges faced by dissenting employees, internal resistance to Project Nimbus continues to grow within Google. The solidarity displayed by employees like Hatfield reflects a broader sentiment of discontent among workers who perceive their labor being exploited to perpetuate oppressive systems. No Tech for Apartheid members remain resolute in their commitment to challenging Google’s complicity in facilitating harmful practices, underscoring a collective determination to advocate for ethical accountability within the tech industry.
