By Lina Irawan 
Caitlin Kalinowski, a highly respected figure who had been at the helm of OpenAI’s critical hardware and robotic engineering teams since November 2023, has announced her departure from the company. Her resignation, conveyed through public posts on X (formerly Twitter) and LinkedIn, casts a stark spotlight on the escalating ethical dilemmas facing the artificial intelligence industry, particularly concerning the integration of advanced AI with military applications. Kalinowski’s exit marks a significant loss for OpenAI in a strategically vital area and serves as a potent symbol of the deep moral divisions emerging within the AI community.
In her candid statement, Kalinowski articulated her reasons for leaving, emphasizing that her decision was driven by "principle, not people." She expressed profound respect for OpenAI CEO Sam Altman and the teams she worked with, underscoring her pride in their accomplishments. However, her core disagreement centered on fundamental ethical lines she believes were inadequately deliberated by the company. Specifically, she cited concerns over "surveillance of Americans without judicial oversight and lethal autonomy without human authorization." These two issues represent cornerstones of the broader debate around AI safety and human rights, striking at the heart of how powerful AI technologies should be deployed in national security contexts. For many ethicists and safety advocates, the prospect of AI systems making life-or-death decisions or enabling widespread, unmonitored surveillance crosses a dangerous threshold that demands rigorous public and judicial scrutiny before any implementation.
Kalinowski’s departure unfolds against a backdrop of intensifying controversy within the AI sector regarding its role in military applications. For years, many leading AI companies and researchers maintained a cautious stance, often explicitly prohibiting the use of their technologies for weaponry or surveillance. OpenAI itself, in its original founding charter, emphasized a commitment to developing beneficial AI that avoids harm, and its early usage policies included a prohibition on military applications. However, this stance has evolved as the company transitioned from a pure non-profit to a "capped-profit" entity, driven by commercial pressures and the immense resources required to develop cutting-edge AI. The quiet removal of the "no military use" clause from OpenAI’s usage policies earlier in 2024 signaled a significant pivot, opening the door to engagement with defense agencies. This shift ignited internal debate and external criticism, laying the groundwork for the ethical friction that has now culminated in Kalinowski’s resignation.
The immediate catalyst for this latest wave of dissent appears to be the recent developments surrounding AI companies and the U.S. Pentagon. Just prior to OpenAI’s agreement, rival AI firm Anthropic, co-founded by former OpenAI safety researchers Dario and Daniela Amodei, reportedly saw its negotiations with the Pentagon collapse. Anthropic, known for its strong emphasis on AI safety and constitutional AI principles, reportedly pushed for stringent limitations on the use of its technology for domestic surveillance and autonomous weapons. Their refusal to compromise on these ethical "red lines" set a powerful precedent, highlighting a principled stand against potential misuse.
Soon after Anthropic’s negotiations faltered, OpenAI reached its own agreement with the Defense Department. This deal reportedly involves deploying OpenAI’s advanced models on a classified government network, signaling a deeper and more direct involvement with military operations than previously imagined. The timing of this agreement, following closely on the heels of Anthropic’s principled refusal, drew immediate criticism. Many employees and external observers characterized OpenAI’s move as "opportunistic," suggesting the company stepped in to fill a void left by Anthropic’s ethical stance. OpenAI CEO Sam Altman himself later acknowledged that the deal’s rollout "looked opportunistic," indicating an awareness of the public relations challenge and the internal discomfort it generated. In an attempt to mitigate the backlash, the company has since moved to clarify specific restrictions on how its systems can be used by the military, asserting adherence to certain ethical boundaries.
An OpenAI spokesperson confirmed Kalinowski’s departure and provided a statement reiterating the company’s position. The spokesperson stated, "We believe our agreement with the Pentagon creates a workable path for responsible national security uses of AI while making clear our red lines: no domestic surveillance and no autonomous weapons. We recognize that people have strong views about these issues and we will continue to engage in discussion with employees, government, civil society and communities around the world." While OpenAI claims to have established clear "red lines," the practicality and enforceability of these limitations in a classified military environment remain a point of contention. Defining "autonomous weapons" is itself a complex and evolving challenge, with various interpretations ranging from fully autonomous systems to those with "human-on-the-loop" or "human-in-the-loop" oversight. Furthermore, ensuring that models deployed on classified networks are not adapted or repurposed for prohibited uses, such as surveillance without judicial oversight, presents a significant oversight challenge for a civilian company. The inherent opaqueness of military operations often clashes with the transparency required for effective ethical governance of AI.
Kalinowski’s professional background underscores the magnitude of her loss for OpenAI. Before joining the cutting-edge AI lab, she had already established herself as a titan in hardware engineering and product development. Her career trajectory demonstrates a consistent ability to lead complex, pioneering projects at the forefront of technological innovation. For nearly six years at Apple, she played a crucial role in designing iconic MacBooks, including the Pro and Air models, honing her skills in precision engineering and user-centric design. She then spent over nine years at Oculus, Meta’s virtual reality subsidiary, where she was instrumental in developing the early generations of VR headsets. This extensive experience positioned her as a leading expert in immersive hardware, a field that demands both visionary thinking and meticulous execution.
Most recently, prior to her move to OpenAI, Kalinowski served as a hardware executive at Meta for approximately two and a half years. There, she spearheaded the ambitious "Orion" project, previously codenamed Project Nazare, which Meta touted as "the most advanced pair of AR glasses ever made." Under her leadership, Meta unveiled a prototype of these augmented reality glasses in September, showcasing her capability to deliver on highly complex, futuristic hardware initiatives. Her recruitment by OpenAI in November 2023 was a clear signal of the company’s serious intent to expand beyond purely software-based AI into the realm of embodied AI and robotics, a field where physical manifestation and interaction with the real world are paramount. Kalinowski’s expertise was precisely what OpenAI needed to bridge the gap between abstract AI models and their practical application in robotics. Her departure therefore represents a significant blow to OpenAI’s ambitions in this critical area, potentially delaying or complicating their efforts to develop advanced robotic systems capable of performing tasks in the physical world, a key step towards achieving general artificial intelligence.
Kalinowski’s principled resignation is not an isolated incident but rather indicative of a broader and growing ethical exodus within the AI industry. Over the past few years, several high-profile AI ethicists, safety researchers, and engineers have left leading AI labs like Google DeepMind and even OpenAI itself, citing similar concerns about the rapid pursuit of commercialization and deployment over rigorous safety protocols and ethical considerations. The perceived sidelining of figures like Ilya Sutskever and the departure of Jan Leike from OpenAI’s Superalignment team earlier this year also pointed to internal tensions regarding the company’s direction and commitment to safety. This "brain drain" of ethically minded talent poses a significant challenge for AI companies, as it erodes public trust, fuels skepticism, and potentially hinders the development of truly beneficial and safe AI systems.
The underlying tension is further exacerbated by the intense geopolitical race for AI dominance. Governments worldwide, particularly the U.S. and China, view advanced AI as a critical component of national security and economic competitiveness. This perceived imperative often creates immense pressure on leading AI companies to engage with defense contracts, offering substantial funding, access to data, and a platform for large-scale deployment. However, this pressure frequently comes with ethical compromises, forcing companies to balance innovation, national security interests, and their stated commitments to responsible AI development. The military-industrial complex’s historical appetite for cutting-edge technology inevitably draws in the most advanced AI capabilities, creating a complex web of ethical, economic, and geopolitical considerations that are challenging for individual companies, let alone the broader society, to navigate responsibly.
Kalinowski’s departure will undoubtedly have ramifications for OpenAI, particularly for its nascent robotics division, which will now need to find new leadership for its hardware engineering efforts. More broadly, it will intensify scrutiny on OpenAI’s military contracts and its stated "red lines," prompting further questions about the enforceability and sincerity of its ethical commitments. For the wider AI industry, Kalinowski’s principled stand may embolden other researchers and engineers to voice their concerns, fostering a more transparent and robust debate about AI’s military applications. It also underscores the urgent need for policymakers to establish comprehensive, internationally agreed-upon frameworks for the ethical development and deployment of AI in warfare, moving beyond self-regulation by tech companies. Ultimately, this incident highlights the ongoing struggle to reconcile rapid technological advancement with fundamental ethical principles, a challenge that will define the future of artificial intelligence.