The Kenyan Parliament has a central role in ensuring that Artificial Intelligence (AI) does not outpace the country’s laws and institutions responsible for protecting citizens, the Communications, Information and Innovation committee (CII) chairperson John Kiarie has said.
Mr Kiarie wants stronger parliamentary oversight as Kenya moves to embrace AI, saying technological innovation must go hand in hand with accountability, protection of human rights and safeguards for citizens.
“Parliament must carefully examine how AI is being developed, procured and deployed, particularly where automated systems can influence decisions affecting people’s livelihoods and access to essential services”, the Dagoreti South MP said.
“AI is likely to cause disruptions to employment, education, healthcare, financial services, social protection, security and justice.”
Mr Kiarie made the remarks when he delivered a keynote address during a webinar on “Synchronising AI Innovations: Legislative Interventions Towards Harnessing AI in Africa.”
Dagoretti South MP John Kiarie, who is the chairperson of the National Assembly's Committee on Communication, Information and Innovation.
Photo credit: Dennis Onsongo | Nation Media Group
The webinar was convened by the Centre for Parliamentary Studies and Training (CPST) and moderated by the director general for Parliamentary Joint Services Clement Nyandiere where experts from Parliament, government and the technology sector examined how Kenya can harness AI while protecting citizens from its potential risks.
The session focused on the need for Parliament to determine whether existing laws are sufficient to deal with emerging AI-related challenges or whether new legislation and safeguards are required.
Grace Githaiga, chief executive and convenor of the Kenya ICT Action Network (KICTANet) said Kenya should not have to choose between technological innovation and human rights.
Dr Githaiga said the two can advance together if the country builds a governance framework that encourages innovation while giving citizens, businesses and public institutions confidence that AI systems are being used responsibly.
“The concern is no longer simply about what personal information is collected and stored. AI systems can analyse large amounts of information, draw conclusions about individuals, create profiles and make or influence decisions based on data whose original purpose may have been very different,” Dr Githaiga said.
She said the new technology has raised difficult questions about the data used to train AI models, the information systems they infer about individuals, and the consequences of automated decision-making.
Participants observed that biases could have serious consequences in areas including recruitment, lending, education, healthcare, insurance and access to public services.
ICT Authority Chief Executive Officer Jessy Maruti called for stronger safeguards as government agencies increasingly turn to AI to improve service delivery.
Mr Maruti said the government cannot escape responsibility simply because a decision has been assisted or made by an algorithm.
“Government accountability cannot be delegated to a vendor or an algorithm. The infrastructure, standards, institutions and accountability are key pillars for responsible AI adoption in government,” Maruti said.
The webinar observed that Kenya’s digital transformation must go beyond internet connectivity to include reliable electricity, fibre networks, cloud infrastructure, secure computing capacity, trusted data systems, cybersecurity and adequate digital skills.
Despite about 96 percent of Kenyans are estimated to live within mobile broadband coverage, a significant usage gap remains, demonstrating that access to a network does not automatically translate into meaningful digital inclusion.
Mr Maruti emphasised the importance of human oversight, especially when AI systems are used in decisions that can affect people’s rights, benefits, safety or access to government services.
He called for clear mechanisms through which citizens can challenge decisions, seek human review and obtain appropriate remedies when AI systems make mistakes.
“Cybersecurity must be considered throughout the life of an AI system, from the collection of data and development of the technology to deployment, monitoring and eventual retirement,” the ICTA boss said.
“There is a need for a risk-based approach to AI regulation, with low-risk applications facing basic safeguards while systems capable of significantly affecting people’s rights or safety would face much stricter requirements.”
Mr Maruti said high-risk systems could require impact assessments, independent review, external validation, registration, regular audits and clear appeal mechanisms.
AI agents, software systems that use artificial intelligence to pursue a goal, make decisions, and execute multi-step tasks independently without constant human supervision.
Photo credit: Shutterstock
Mr Kiarie said in the case of Parliament, the emerging AI landscape presents five major responsibilities in areas of legislation, oversight, representation, resourcing, and review.
He said his committee is expected to spearhead any necessary policy framework and continually serve as an important platform for examining the issues, including the way government agencies procure AI technologies and whether taxpayers are receiving value while citizens’ rights are protected.
Mr Kiarie said his committee is also expected to ensure that communities affected by AI have an opportunity to participate in shaping the rules governing the technology, while regulators and oversight institutions receive the technical and financial resources necessary to do their work.
The web meeting proposed a government roadmap that includes taking stock of existing AI systems and pilots, introducing interim rules for AI use and procurement, identifying accountable officers and developing risk classifications and impact assessments.
High-risk systems could subsequently be subjected to certification, interoperability testing and publication of performance and audit findings.
The meeting concluded that Kenya should not slow down innovation, but neither should it allow technology to move ahead without safeguards.
Mr Kiarie said the challenge facing Kenya is not how quickly the country can adopt AI, but how responsibly it can do so while protecting human dignity, equality, privacy, freedom of expression, access to information and accountability.
