Government Surveys Lack Safeguards Against Bot Responses

A study finds Ottawa has done little to confirm its decisions rest on surveys of real people rather than AI-driven bots.

Updated on Nov. 17, 2026 in Artificial Intelligence

Bold flat-color editorial illustration of stacked ballot-box forms with a cracked seam and a wedged gear on a neutral plinth, evoking doubt about the integrity of survey data.
Ottawa has done little to confirm that federal decisions rest on survey responses from real Canadians rather than AI-driven bots, a new study found. AI Illustration. Upload story photo >

Live Poll

Do you trust federal government surveys to reflect the views of real Canadians?

A study found the federal government has done little to ensure its decisions are based on surveys of real people rather than bots. Bots can mimic human respondents by matching reading times to a given education level, generating human-like mouse movements and inserting believable typos.

Why it matters

Government policy built on survey data depends on responses from actual Canadians, so unverified respondents could distort the evidence used in federal decisions. Public Services and Procurement Canada has said it is incorporating additional guardrails against the use of AI in government research.

The evidence is a study of federal survey practices, which found the government has done little to verify respondents. Phone surveys are costlier than online surveys and reach young people less effectively, while the extent of bot-generated responses in federal data remains unquantified.

The players

Public Services and Procurement Canada

The federal department responsible for purchasing and common services for the government, which has updated its research guidance in response to AI-related risks.

Abacus Data

An Ottawa-based polling firm that employs quality control checks and flags where AI is being used in its survey work.

The details

Abacus Data, an Ottawa polling firm, uses quality control checks and flags where AI is used, and works with the PureSpectrum platform, which assigns quality control scores to responses from online panels. Public Services and Procurement Canada has updated its research guidance, strengthened statements of work and research proposals, and increased data quality measures.

Timeline

  1. November 17, 2026: The article reporting the study's findings was published.

The Tech Race

The study's findings place Canada's federal survey practices within a wider contest over whether AI-generated respondents can pass as people in public research. Ottawa's move to write higher data standards into research contracts reflects how government procurement is becoming a tool for setting those standards.

Canadians who take part in online surveys may be asked for responses that government researchers later rely on, so the integrity of those panels affects the policies that follow. Government decisions shaped by bot-generated data could change the services and programs residents receive.

The takeaway

Survey data shapes federal decisions only when the people answering are real, so verifying respondents matters as much as the questions asked. Canadians can expect more stringent research contracts as government works to close these gaps.

Further reading

For related coverage, explore more Artificial Intelligence stories on how AI is affecting public institutions in Canada.

Source note: This article includes information reported by Ottawa Citizen.

Live Poll

Do you trust federal government surveys to reflect the views of real Canadians?