The development of AI has numerous implications for the functioning of parliaments:
• For parliamentarians, AI can help improve the conditions under which they carry out their day-to-day role as legislators (assisting with the drafting of legislation and amendments) as well as their role in monitoring the government (processing public data, analysing and summarising reports);
• For parliamentary administrations, AI can help modernise departments in terms of the efficiency, accessibility and transparency of parliamentary work (transcription, translation, subtitling and summarising of debates; processing legislative texts and amendments; documentary research; strengthening cybersecurity).
These opportunities for modernising Parliaments can also help to rebuild connections with citizens allowing them to become more closely involved in the political process through new tools for citizen participation.
Nonetheless, AI systems may also pose risks to the proper functioning of Parliaments:
• From a democratic perspective, AI can facilitate the spread of false or misleading information, foreign interference, the polarisation of public debate and online violence, particularly gender-based violence. Its use in Parliaments can also weaken the role of human responsibility in political decision-making, can undermine the integrity of public debate and can erode the trust of citizens in Parliament.
• Within Parliaments themselves, several issues must be managed: algorithmic biases that exacerbate existing inequalities and discrimination; the protection of sensitive data; cybersecurity; strategic dependence on foreign private or state actors; a lack of transparency in decision-making processes as well as transformations in certain professions.
Against this backdrop, the Speakers of the G7 Assemblies are called upon to consider and implement measures that ensure the responsible and controlled use of AI in parliamentary work in a manner consistent with democratic values :
• Do your Parliaments have an AI strategy or policy framework in place? Does it apply to parliamentarians? To staff? To administrative services?
• How can we support the informed and responsible use of AI by parliamentarians, particularly with regard to data confidentiality, the risks of algorithmic bias and foreign interference?
• What initiatives and pilot schemes have already been implemented in your Assemblies to improve day-to-day legislative work through the use of AI (transcription, translation, cybersecurity, documentary research, legislative functions such as drafting bills and amendments)?
• How can we guarantee transparency for citizens and ensure that political accountability always remains with human decision-makers when AI tools are used in parliamentary work?
• Are there any uses of AI that should be prohibited in parliamentary work?
• How can parliamentary diplomacy make a meaningful contribution to the debate on AI? Would an initiative at the level of the G7 Parliamentary Group be worthwhile?
