AI, Democracy, and the Question of Power
STATEMENT OF THE EUROPEAN LEFT
AI must serve people and democracy — not reinforce concentrated power.
Brussels, 3 March 2026. The European Left expresses deep concern about the changing role of major artificial intelligence developers in military and defense systems. These developments raise important questions about democratic oversight, civil liberties, and global stability.
Recent news shows that OpenAI, one of the most popular AI developers, signed a contract with the U.S. Department of Defense. This contract allows its advanced language models to be used on classified military networks and domestic surveillance.
After facing public criticism, OpenAI CEO Sam Altman announced changes to the agreement. These changes aim to clarify protections against domestic surveillance and to exclude certain intelligence agencies unless additional terms are agreed upon.
This deal came into effect following the collapse of similar negotiations with rival firm Anthropic, which had reportedly resisted Pentagon terms that lacked clear prohibitions on mass domestic surveillance and fully autonomous weapon systems — systems capable of making lethal decisions without meaningful human oversight. President Donald Trump’s administration subsequently directed federal agencies to end the use of Anthropic’s technology.
The European Left welcomes the public outcry and open debate that have emerged around these issues. The strong reaction from citizens, including campaigns for ethical AI governance and responsible corporate behavior, such as QuitGPT, shows a growing global concern about the direction of cutting-edge AI systems and their integration into government power and military structures.
At the same time, the European Left remains seriously worried about fundamental risks:
- Militarization of AI: Using commercial AI systems for military or defense purposes without strong democratic oversight can entrench technologies that escalate conflict and diminish human control over life-and-death decisions.
- Inadequate Safeguards: Relying on existing legal limits instead of strong, clear prohibitions leaves open the possibility that AI could still be used for large-scale data analysis, profiling, or surveillance that is technically legal under current rules.
- Corporate-State Entanglement: Close ties between powerful AI firms and state security agencies without enforceable democratic safeguards weaken public trust and accountability.
- Erosion of Civil Liberties: Advocates for civil liberties warn that even well-meaning legal limits might not prevent invasive data use, especially if current surveillance authorities are interpreted broadly in practice.
Thus, the European Left calls for clear and urgent action:
- A global ban on fully autonomous weapons—machines should never be allowed to decide who lives and who dies.
- Full transparency and democratic oversight whenever AI is used by military or security authorities.
- Strong, enforceable protections to prevent mass surveillance and invasive data profiling.
- Public investment in AI systems that are governed democratically, open, and designed to serve the public interest—not private profit.
- Meaningful citizen participation in shaping how AI is developed, regulated, and deployed.
Artificial intelligence will significantly impact our economies, democracies, and daily lives. It must be guided by human rights, peace, and social justice—not by narrow commercial interests or geopolitical competition.
Citizens are not powerless. We can decide which AI tools we use and support.
By choosing services that respect democratic values and human rights, we send a clear message about the kind of technology future we want.