Meta's AI ad delivery skews New Zealand political ads by gender, analysis finds

Researcher calls for transparency as algorithmic optimization, not party strategy, appears to shape who sees election ads

edit
By LineZotpaper
Published
Read Time3 min
An analysis by Radio New Zealand has found that many political ads run on Meta's platforms in New Zealand reached almost exclusively men—or, for the Green Party, exclusively women—prompting concerns that AI-driven ad delivery, rather than deliberate party strategy, is deciding who sees political messaging. Researchers say the findings highlight an urgent need for transparency in algorithmic political advertising.

Radio New Zealand's analysis of political advertising on Facebook and Instagram has revealed sharp gender skews in who saw the ads. Of more than 500 ads run by the National Party this year, 92 were seen only by men, attracting up to 3.9 million views and costing as much as $21,000. Only one ad was seen solely by women—for a single day, at a cost of less than $100—and the party says it did not place it. ACT showed a similar tilt, with 14 ads seen only by men and none solely by women. For the Greens, the pattern was reversed: 13 ads were seen only by women and none only by men.

At first glance, these patterns might suggest parties are deliberately targeting separate audiences. But Katherine Jensen, a professional teaching fellow at the University of Auckland Business School, argues that assumption is outdated. For much of the past decade, digital advertising allowed advertisers to select audiences based on age, gender, or interests, and platforms delivered accordingly. Meta has since shifted to AI-driven optimisation. Its Andromeda system narrows tens of millions of potential ads and variations to a few thousand likely matches before ranking them, and its Advantage+ tools can automate decisions about audiences, budgets, and placement.

This makes the distinction between targeting and optimisation crucial, Jensen writes. Advertisers choose who they initially want to reach, while the platform learns who responds and increasingly delivers ads to them. A party can begin with a deliberately broad audience and still end up reaching disproportionately more men, younger voters, or people already inclined to agree with its message.

Researchers have documented this problem for years. A 2019 study found Facebook's ad delivery could skew audiences by gender and race even when advertisers used inclusive targeting. Other research found voters statistically less likely to support a party could cost two or three times as much to reach, while campaigns had little visibility into how those decisions were made. Similar pricing effects have appeared in elections abroad, including in India, where Facebook charged the ruling BJP lower rates for ads than other parties during elections.

Meta presents its automated systems as improving relevance and campaign performance, and it maintains a public Ad Library for political ads. But Jensen argues the available data is insufficient. She says New Zealand urgently needs transparency rules so that both parties and voters understand how algorithmic systems shape political messaging. Without such rules, she warns, the people who see political ads—and those who do not—may be chosen by algorithms rather than by democratic deliberation.

§

Analysis

Why This Matters

  • Political advertising is a core part of democratic campaigns; if AI systems determine who sees ads, voters may receive incomplete or biased information about parties and policies.
  • New Zealand lacks specific transparency rules for algorithmic ad delivery, leaving a gap in electoral oversight as AI-driven tools become standard.
  • The skew is not just about gender—it could affect other demographic groups, potentially shaping election outcomes in ways neither parties nor voters understand.

Background

  • For most of the past decade, digital political advertising on Meta platforms worked through manual audience selection: advertisers chose age, gender, interests, and platforms delivered ads accordingly.
  • Meta has since introduced AI-driven systems, including the Andromeda retrieval engine and Advantage+ campaign tools, which automatically optimize audience selection and delivery based on predicted user response.
  • Researchers have flagged the risk since at least 2019, when a study showed Facebook's delivery algorithms could skew by gender and race even with inclusive targeting. The RNZ analysis of New Zealand's 2026 election campaign is one of the first real-world examinations of this effect in a live political contest.

Key Perspectives

Researchers: Katherine Jensen and others argue the opacity of Meta's algorithms undermines electoral integrity and call for regulatory transparency, such as requiring platforms to disclose delivery metrics and algorithmic logic. Political parties: National and ACT say they did not deliberately target men, and the Greens similarly did not intend to reach only women. They may benefit from reaching persuadable voters, but they have limited visibility into how Meta's optimization changes their intended audience. Meta: The company emphasizes that Advantage+ improves campaign performance by showing ads to users most likely to engage, and it provides an Ad Library for transparency. Critics, however, argue the library's aggregated data does not reveal the underlying algorithmic decisions, leaving both parties and regulators in the dark.

What to Watch

  • Whether Meta responds by releasing more granular delivery data for political ads in New Zealand.
  • Any moves by the New Zealand Electoral Commission or Parliament to update advertising rules to cover algorithmic targeting.
  • Further journalistic and academic audits of the Meta Ad Library in the lead-up to the next election.

Sources

newspaper

Zotpaper

Articles published under the Zotpaper byline are synthesized from multiple source publications by our AI editor and reviewed by our editorial process. Each story combines reporting from credible outlets to give readers a balanced, comprehensive view.