Build Canada Created a Robot MP. It's Not Cute Shades of 'Big Tech authoritarianism' A robot programmed for 'economic freedom' What does 'build Canada' really mean?
Jun 26, 2026•11 min•Ep. 43
Listen to this episode in Metacast mobile app
Don't just listen to podcasts. Learn from them with transcripts, summaries, and chapters for every episode. Skim, search, and bookmark insights. Learn more
Episode description
Tech bros made an AI 'politician' to love 'economic freedom.' Who are its constituents? Last in a series. … Article written by Christopher Holcroft.
Voting for fellow community members to represent our interests in Parliament is a very human activity here in Canada. So is consenting to have that elected Parliament legislate on our behalf. Human, too, is having something to say publicly about proposed legislation.
These are, in fact, fundamental rights in one of the oldest continuing democracies in the world.
We should therefore be wary of degrading gimmicks such as the recording of made-up votes by a phoney member of Parliament. Although that is not discouraging a group of wealthy tech bros from trying.
Shortly after last year's election, the advocacy group Build Canada created what it calls the world's "first AI member of Parliament." The "Builder MP" project uses OpenAI's GPT-5 to review every proposed piece of legislation introduced in Parliament, summarize it, analyze it against a specific set of criteria consistent with Build Canada's economic freedom-focused policy agenda, and assign it a vote (yes, no or abstain).
The AI MP also proposes "question period"-style questions for some bills. Sometimes, it suggests potential improvements to a bill.
OpenAI is the Sam Altman-led company that Canada's privacy commissioner recently said violated federal and provincial privacy laws in training its ChatGPT tool.
This is also the company being sued by families in Tumbler Ridge, B.C., for failing to alert authorities it had knowledge of a planned mass shooting before it took place. After months of waiting, the Liberal government chose not to include a mandatory reporting requirement in its recently announced legislation on online harms, disappointing B.C. Premier David Eby.
At the time of the AI MP launch, Build Canada CEO Lucy Hargreaves sought to frame the tool the same way the tech industry usually frames AI use — as an inevitability. "This project is about showing that the same technologies transforming industries can also modernize our public life.... AI will reshape how societies function, including how democracies govern themselves," she wrote.
Such framing foreshadows a very real potential danger consistent with the rise of what Azadeh Akbari, professor of critical data and technology studies at Goethe University, refers to as "Big Tech authoritarianism": the offloading of political rights, civic responsibility and democratic accountability from citizens and their elected leaders to private interests and their programmable machines.
Presently, real democracies are experimenting with integrating AI machines into their public institutions and political decision-making bodies. Albania has appointed an AI cabinet minister. The prime minister of Sweden regularly consults AI for a second opinion. The United Kingdom is using AI to advise on local planning decisions. The European Union is assessing the laws of prospective member countries for compatibility with the union through AI.
In Canada, public submissions to Parliament as part of the policy consultation on a national AI strategy were reviewed and summarized by AI, a role normally reserved for non-partisan public servants. Canada's AI minister, Evan Solomon, admitted to turning to AI to receive a briefing on a bill, a process that reportedly proved badly flawed.
Since the new Liberal government came to power, the AI MP has reviewed and opined on about 200 bills. The results have been underwhelming.
For example, on Bill C-253, "An Act to develop a national framework for a guaranteed livable basic income," the AI MP gets a basic fact wrong. Whereas the actual bill proposes "a national framework for the implementation of a guaranteed livable basic income program throughout Canada for any person over the age of 17," the AI MP claims the bill proposes a basic income be delivered to "all people in Canada aged 17 and over." Misidentifying the age of applicability is a significant, ...
Voting for fellow community members to represent our interests in Parliament is a very human activity here in Canada. So is consenting to have that elected Parliament legislate on our behalf. Human, too, is having something to say publicly about proposed legislation.
These are, in fact, fundamental rights in one of the oldest continuing democracies in the world.
We should therefore be wary of degrading gimmicks such as the recording of made-up votes by a phoney member of Parliament. Although that is not discouraging a group of wealthy tech bros from trying.
Shortly after last year's election, the advocacy group Build Canada created what it calls the world's "first AI member of Parliament." The "Builder MP" project uses OpenAI's GPT-5 to review every proposed piece of legislation introduced in Parliament, summarize it, analyze it against a specific set of criteria consistent with Build Canada's economic freedom-focused policy agenda, and assign it a vote (yes, no or abstain).
The AI MP also proposes "question period"-style questions for some bills. Sometimes, it suggests potential improvements to a bill.
OpenAI is the Sam Altman-led company that Canada's privacy commissioner recently said violated federal and provincial privacy laws in training its ChatGPT tool.
This is also the company being sued by families in Tumbler Ridge, B.C., for failing to alert authorities it had knowledge of a planned mass shooting before it took place. After months of waiting, the Liberal government chose not to include a mandatory reporting requirement in its recently announced legislation on online harms, disappointing B.C. Premier David Eby.
At the time of the AI MP launch, Build Canada CEO Lucy Hargreaves sought to frame the tool the same way the tech industry usually frames AI use — as an inevitability. "This project is about showing that the same technologies transforming industries can also modernize our public life.... AI will reshape how societies function, including how democracies govern themselves," she wrote.
Such framing foreshadows a very real potential danger consistent with the rise of what Azadeh Akbari, professor of critical data and technology studies at Goethe University, refers to as "Big Tech authoritarianism": the offloading of political rights, civic responsibility and democratic accountability from citizens and their elected leaders to private interests and their programmable machines.
Presently, real democracies are experimenting with integrating AI machines into their public institutions and political decision-making bodies. Albania has appointed an AI cabinet minister. The prime minister of Sweden regularly consults AI for a second opinion. The United Kingdom is using AI to advise on local planning decisions. The European Union is assessing the laws of prospective member countries for compatibility with the union through AI.
In Canada, public submissions to Parliament as part of the policy consultation on a national AI strategy were reviewed and summarized by AI, a role normally reserved for non-partisan public servants. Canada's AI minister, Evan Solomon, admitted to turning to AI to receive a briefing on a bill, a process that reportedly proved badly flawed.
Since the new Liberal government came to power, the AI MP has reviewed and opined on about 200 bills. The results have been underwhelming.
For example, on Bill C-253, "An Act to develop a national framework for a guaranteed livable basic income," the AI MP gets a basic fact wrong. Whereas the actual bill proposes "a national framework for the implementation of a guaranteed livable basic income program throughout Canada for any person over the age of 17," the AI MP claims the bill proposes a basic income be delivered to "all people in Canada aged 17 and over." Misidentifying the age of applicability is a significant, ...
For the best experience, listen in Metacast app for iOS or Android