As AI moves from experimentation to critical infrastructure, trust is emerging as the defining challenge of its next phase. In this session, using the Mythos episode as a wake-up call, Christel Heydemann will explore why AI must now be seen not only as a technological breakthrough, but also as a matter of strategic autonomy, security, economic control and human responsibility. Drawing on Orange’s role as a trusted operator of networks, cloud, cybersecurity and digital infrastructure, she will share a clear conviction: in the age of agentic AI, dependency and loss of control are the new vulnerabilities. A strategic perspective on how Europe can accelerate in AI without giving up control.