Public Sector Intelligence Report: May to October 2026

Face Forward Intelligence ReportsSix-month report · May 1 to October 31, 2026
1.48M

Face Forward comments analysed in six months, May 1 to October 31, 2026.

Report · Public sector

Six months in, abuse is the working environment for women in public life

The Areto Public Sector Intelligence Report, May 1 to October 31, 2026. Face Forward's first six months and a first comparison with men in public life.

By Areto Team · Delivered in research partnership with the ANU Global Institute for Women's Leadership

Between May 1 and October 31, 2026, Face Forward analysed 1,483,343 comments on the social media accounts of women and gender-diverse people in public life across Australia, New Zealand and the Pacific. Areto automatically removed 693,181 of them (46.7%) before participants had to see them. At 60 seconds of manual review per removed comment, that's about 11,553 hours of staff time saved.

The bigger finding is the shape of the abuse. Close to half of all comments have been removed as harmful in every month since July, with no single flashpoint behind any of them. October, the sixth month, looked the same.

Face Forward is a fully-funded initiative, giving free, real-time protection to women and gender-diverse people in public life. It's delivered in research partnership with the ANU Global Institute for Women's Leadership (GIWL). This report reads its first six months together and adds a first comparison with men in public life protected through Areto's paid service.

Key findings: Face Forward, six months

  • Close to half of comments were removed as harmful every month from July to October (48.4%, 47.7%, 47.1% and 46.0%). In October, 44.8% of comments were abusive at the low threshold.
  • Facebook carried more than nine in ten abusive comments in every month. In October, it also had the highest abuse rate of the major platforms (46.2%).
  • Gendered abuse leads the identity-based categories. In October, sexist abuse and gender microaggressions (12,850) outweighed racist and xenophobic abuse combined (8,448).
  • More than half of comments were negative every month from July. Participants received about 835,000 negative comments over six months. Areto removes abuse, not criticism.

Key findings: women and men in public life (directional)

  • Women's accounts saw a higher share of comments removed every month from June, by 13.2 to 23.6 points (October: 46.0% against 32.8%).
  • Same racism, plus an added layer. In October, per 1,000 comments, Face Forward accounts received 6.1 times the sexist abuse men's accounts did, 4.0 times the gender microaggressions and 3.4 times the transphobic abuse. Racist abuse arrived at the same rate.
  • The daily grind is shared. Personal attacks were 83.9% of abusive comments for both groups in October.

Part 1. Face Forward, May to October 2026

Six months of protection for women and gender-diverse people in public life across Australia, New Zealand and the Pacific. Funded participants only.

Six months at a glance

Volume grew from May to September, mostly because more participants joined, and eased 4% in October.

MonthCommentsRemovedRemoved shareHours saved*
May9,1181,43915.8%24
June51,88919,75038.1%329
July324,564157,00348.4%2,617
August321,173153,14247.7%2,552
September397,193187,20947.1%3,120
October379,406174,63846.0%2,911
Six months1,483,343693,18146.7%11,553

Restated figures from the October 2026 report, funded participants only. Earlier monthly reports also included men in public life on Areto's paid service, so these figures are lower than first published (July: 324,564 comments against 352,244). The men's accounts are reported separately in Part 2. * At 60 seconds of manual review per removed comment.

Abuse settled high

Share of comments removed as harmful

May
15.8%
June
38.1%
July
48.4%
August
47.7%
September
47.1%
October
46.0%

Removed comments as a share of comments analysed, funded participants only. Bars run from 0 to 60%.

May set a starting point of 15.8%. In June, the share rose to 38.1%.

From July the shape changed. Participation in Face Forward grew, as did the percentage of comments flagged for moderation. The share of comments removed has stayed between 46.0% and 48.4% for four months running, with no single post or event behind it. October's abuse rate was 44.8% at the low threshold: nearly half of all comments participants received were abusive.

Abuse on these accounts isn't an event to wait out. It's the persistent working environment.

Averages hide the heaviest loads

The monthly rate is an average across every participating account, and some people carry far more. In October's sample, five accounts received nearly 60% of all abusive comments. Of the 45 accounts with substantial volume in the sample, 31 had abuse rates above 25% and 14 had more than half their comments flagged as abusive. The August and September reports found the same pattern, with the most heavily targeted accounts at 50% to 72%. That's the case for protecting people one account at a time, not only reporting averages.

Facebook carries the abuse

Every monthly report shows the same Facebook dominance, with more than nine in ten abusive comments there in every month from May to October. The earlier reports also show Instagram's abuse rate doubling between May and August, from 11.6% to 35.2%. In June, one participant faced 58% abuse on Facebook and 0% on Instagram in the same week. The platform gap can be personal.

FacebookShare of commentsShare of abusive commentsAbuse rate
May85.3%91.5%23.0%
June92.2%95.9%40.1%
July95.0%96.7%47.1%
August91.9%94.4%45.4%
September89.2%93.1%43.9%
October90.4%93.2%46.2%

May to September as first published in the monthly reports, which also included some men in public life on Areto's paid service. October: funded participants only.

In October, Facebook carried 90.4% of comments and 93.2% of abusive comments. Its abuse rate of 46.2% is the highest of the major platforms. Instagram (34.4%) and TikTok (20.0%) are less hostile, but far from safe.

October 2026CommentsAbusiveAbuse rateShare of abuse
Facebook342,921158,51746.2%93.2%
Instagram28,4269,76934.4%5.7%
TikTok7,6161,52320.0%0.9%
YouTube44321047.4%0.1%
All platforms379,406170,01944.8%100%

From the October report's platform table. YouTube's volume is too small to compare reliably.

Gendered abuse leads the identity layer

Personal attacks are the bulk of the abuse, the daily grind of hostile content. Underneath them sits a second layer aimed at who participants are rather than what they said, and it's gendered first.

The monthly reports show how the gendered layer grew. Sexist comments first outnumbered racist comments in July, and the lead widened in August and September. Gendered abuse more than doubled its share of abusive comments over those five months, from 4.9% to 10.8%. Race-based abuse didn't go away: its flags rose from 419 in May to 19,740 in September, while its share fell as the mix of accounts changed. Transphobic abuse grew its share in every month, from 0.4% to 2.5%.

In October, personal attacks appeared in 83.9% of abusive comments. Sexist abuse (7,755) and gender microaggressions (5,095) were the two largest identity-based categories, together outweighing racist and xenophobic abuse combined (8,448).

October 2026FlagsPer 1,000 comments
Personal attack142,579375.8
Sexist7,75520.4
Gender microaggression5,09513.4
Racist4,92313.0
Xenophobic3,5259.3
Ableist3,1058.2
Transphobic2,7797.3
Islamophobic2,5766.8
Body shaming1,2743.4

Selected categories from Areto's October Face Forward report. Categories overlap, so one comment can count in more than one.

More than one in five identity-based comments in the October sample targeted several identities at once, such as ethnicity, religion and gender in the same comment. Islamophobic comments clustered around participants who are visibly from marginalised communities. Transphobic abuse was often used as a weapon against women participants generally, through "define a woman" style attacks, whatever they had posted.

Negative isn't abusive

More comments were negative than were removed, every month. In October, 57.5% of comments were negative and 44.8% were abusive. October's sample shows both sides of that line. Around three in ten negative comments weren't abusive: they carry critique and real political views, sometimes hostile, that moderation is designed to leave up. And one in seven abusive comments wasn't negative in tone at all.

That's the evidence that protection isn't censorship. Areto is built for nuance. It can tell criticism from hate, so participants keep hearing from constituents who disagree with them while the abuse comes down. It also catches abuse that sounds calm: loaded questions such as "what is a woman?", mock politeness and backhanded remarks designed to undermine.

The raw environment barely moved. Positive sentiment fell from 42.6% in May to 8.0% in July and has stayed near 10% since. Negative sentiment has been above half every month from July, and 57.5% in October. Over six months, participants received about 835,000 negative comments. Monthly rates don't show what it's like to read those comments day after day.

In their own words

Participants and their teams describe what changes when the abuse comes down.

Using Areto has given me so much valuable time back. Instead of spending hours each week moderating abusive and offensive content, I can focus on engaging with constituents, listening to community concerns, and delivering on the responsibilities I was elected to fulfill.
MP, Australia · September 2026
Areto has been a game changer for my work and for our page.
Social media manager, Australia · October 2026
It's been really valuable for making sure that, as comments arise, they are flagged and much easier to manage.
Mayor, New Zealand · July 2026
I feel empowered to go back on social media.
MP, Australia · August 2026
I'm very grateful and wish I knew about this sooner. It's all just there and ready to report as evidence.
Councillor, Australia · October 2026
While it is disturbing to me that we have reached a point where we need this protection I am very grateful. I have had my social media on pause as I was becoming too stressed to use it because of the fall out.
Local councillor, New Zealand · May 2026
I'm looking forward to a life of not having to brace myself before opening an app!
Election candidate, New Zealand · June 2026
We have really enjoyed working with Areto to do all we can to help our candidates be safe online. Those that are using it, have reported a marked decline in trolling and nasty comments. This makes candidates' and their social managers' lives so much less miserable.
Campaign manager, New Zealand · June 2026
A great initiative and essential for any inclusive and responsive democracy.
Campaign manager, Fiji · September 2026

Part 2. Women and men in public life

A first comparison with men in public life protected through Areto's paid service. Directional.

Women and men, side by side

Read as directional. Men in public life in Australia and New Zealand (MPs, senators, candidates and party pages) on Areto's paid service, not part of Face Forward. A much smaller group (282,893 comments analysed), not matched by role, party or profile.

Men in public life face hostility too. But on every measure the export allows, women's accounts faced a harsher environment. From June, the share of comments removed as harmful ran 13.2 to 23.6 points higher on Face Forward accounts every month, and negative sentiment ran 7.4 to 14.5 points higher.

Share of comments removedMen in public lifeFace ForwardDifference
June22.9%38.1%+15.2 pts
July24.8%48.4%+23.6 pts
August28.5%47.7%+19.2 pts
September31.7%47.1%+15.4 pts
October32.8%46.0%+13.2 pts

Removed comments ÷ comments analysed (men: from monthly moderation counts). Face Forward: funded participants only. May is left out because both groups were very small that month.

Facebook dominates for men too

Facebook carried 92.2% of comments and 94.1% of abusive comments on men's accounts over six months. Men's accounts have no TikTok. Instagram was harsher for women: in October, 34.4% of comments on Face Forward Instagram accounts were abusive, against 23.4% for men. On Facebook it was 46.2% against 33.7%.

Men's pages host abuse aimed at others

In a sample of September comments on men's accounts, most racist, xenophobic and Islamophobic abuse was aimed at communities rather than the account holder, riding on debates about migration, housing and cost of living. Most gendered abuse on men's pages landed on women in the comment thread. A public figure's page is a public space: moderation protects the constituents reading it, not only the account holder.

Same racism, plus an added layer

October is the one month where both groups can be compared category by category on the same basis. Read per 1,000 comments, so groups of different sizes can be compared, the gendered layer is what separates women's experience from men's. Racist abuse arrived at almost exactly the same rate for both. The difference is aimed at who women are, not what they do.

The daily grind is shared

Personal attacks made up 83.9% of abusive comments on men's accounts in October and 83.9% on Face Forward accounts. Both groups share the same base layer. What changes is how much abuse arrives, and the gendered layer on top.

October 2026, per 1,000 commentsMen in public lifeFace ForwardFace Forward vs men
Personal attack276.3375.81.4×
Sexist3.320.46.1×
Gender microaggression3.313.44.0×
Transphobic2.27.33.4×
Racist13.313.0About the same
Xenophobic6.89.31.4×
Islamophobic2.86.82.4×

From each source's October category counts and comments analysed. Categories overlap, so one comment can count in more than one.

About this data

Where the data comes from. Every figure comes from the Areto platform, which moderates comments in real time. Part 1 draws on the Face Forward Monthly Intelligence Reports, May to October 2026. Part 2 draws on Areto platform data for men in public life on Areto's paid service.

Funded participants only. From October, Face Forward figures cover funded participants only, and the October report restates May to September on that basis. Six-month totals and trends use restated figures from the earlier published Face Forward reports. Where a monthly trend is only available from the earlier reports, which also included some men in public life on Areto's paid service, it's marked as first published.

Share removed and abuse rate. Share removed is removed comments as a share of comments analysed, and it's available for every month. Abuse rate is abusive comments at the low threshold as a share of comments analysed. The October report gives it for October only, so month-by-month comparisons use share removed.

Categories. Categories aren't mutually exclusive, so category totals exceed abusive comments.

Staff time saved. Estimated at 60 seconds of manual review per removed comment.

Sentiment. Measured before moderation, reflecting the raw comment environment.

Observations. Qualitative observations come from a representative sample of comments rather than a full review.

Men in public life. A smaller group, not matched by role, party or profile, so comparisons are directional.

Anonymity. No participant, staffer or commenter is named.

The monthly reports

This report brings together Face Forward's first six monthly Intelligence Reports. Each one is still available on its own.

Next
Next

Face Forward Intelligence Report Release #6