Policy Brief: Building Trusted Chatbots for Critical Thinking - Lessons from the TITAN EU-funded research and innovation project

TITAN Lessons Learned: Building Trusted Chatbots for Critical Thinking is a short policy brief that distills key insights from the EU-funded TITAN research and innovation project. Drawing on three years of technical development, co-creation, piloting, and ethical and legal assessment, the brief synthesizes practical lessons on how AI-driven, Socratic chatbots can be designed to support critical thinking and digital resilience rather than replace human reasoning.

The brief presents a structured set of findings covering communication, transparency, co-creation, technical flexibility, interdisciplinary collaboration, human oversight, and stakeholder expectation management, illustrated through concrete challenges encountered during the project. It highlights why questioning-oriented AI systems require accessible narratives, meaningful transparency, psychologically safe environments, and continuous ethical and legal oversight to build trust and effectiveness.

Targeted at policymakers, educators, AI developers, and public institutions, the brief concludes with a set of actionable policy recommendations and forward-looking implications for future digital literacy and responsible AI initiatives. Together, these lessons position TITAN as a mature prototype and learning case that informs how trustworthy, human-centred AI systems can be responsibly developed and governed in real-world educational and civic contexts.

Note: TITAN has received funding from the EU Horizon Europe research and innovation programme under grant agreement No.101070658, and by UK Research and innovation under the UK governments Horizon funding guarantee grant numbers 10040483 and 10055990.