Regulators Will Soon Use AI To Screen Ad Compliance 24/7...
Author
Stu Sheridan
Date Published

In short:
- In January 2023 the ACCC swept influencer content across six platforms and found that 81% of the posts it reviewed raised concerns under the Australian Consumer Law.
- The sweep covered 118 posts. That figure is the important one.
- Australian advertising enforcement has always operated on samples, because human review has a hard ceiling.
- We're predicting that regulators will soon use automated digital monitoring capabilities.
- When that capability matures, the sample stops being 118 and becomes everything in the category all of the time 24/7 365 days a year... and the low enforcement rate most brands have quietly relied on disappears.
- You as a big or small alcohol, gambling, supplement or other brand in regulated industry need to be awake to this, right now.
What did the ACCC's influencer sweep actually find?
The ACCC's January 2023 internet sweep reviewed influencer content across Instagram, TikTok, Snapchat, YouTube, Facebook and Twitch. It found that 81% of the posts reviewed raised concerns under the Australian Consumer Law for potentially misleading advertising.
Broken down by sector, 96% of fashion influencers reviewed made concerning posts. At the low end, 73% of gaming and technology influencers reviewed made concerning posts.
The most common issue was influencers not disclosing that they had received payment, gifts or other incentives. Other issues included vague disclosure terms such as "sp", "spon", "sponcon", "collab" and "ambassador", disclosures buried at the end of long captions, and disclosures formatted to be hard to notice.
The sweep covered 118 posts.
Why 118 is the number that matters
Australians encounter advertising continuously, across feeds, search, display, retail media, email and out of home. Nobody agrees on the daily figure, but we've heard its somewhere around 4,000 ads a day!
Whatever the real figure is, an ordinary person passes more advertising in a morning than the regulator examined in that entire sweep.
The other ACCC sweeps are the same shape. A separate sweep reviewed 137 businesses for fake or misleading reviews and found 37% had engaged in concerning conduct. A sweep of retail website terms covered just over 2,000 sites.
The ACCC has run advertising sweeps around Black Friday and Boxing Day for two consecutive years and has said it will continue.
These are all human-scale numbers, and that is the point. They are not a measure of how compliant the Australian market is. They are a measure of how much a team of people can review.
Which means the enforcement rate most brands have been implicitly relying on is a statement about regulator capacity, not about their own risk.
Enforcement is still majority, reactive
On 24 March 2026 the ACCC issued two infringement notices over allegedly misleading influencer reviews posted to Instagram by an online retailer. Penalties of $39,600 were paid.
It was the first time the regulator had imposed a financial penalty for failing to disclose paid influencer content, an area that had previously sat with Ad Standards.
The conduct involved more than 100 instances between August 2024 and September 2025 in which the business instructed influencers not to disclose that products had been gifted in exchange for reviews. The gifted items were valued at roughly $50 to $400 each.
The ACCC has stated that payment of a penalty specified in an infringement notice is not an admission of a contravention.
The detail that matters most is how the matter started. The ACCC's investigation began when an influencer reported concerns about a written agreement they had been presented with. Not a sweep. A tip-off.
A further set of infringement notices totalling $138,600 followed in mid-2026 against a separate Australian company in connection with social media content.
So the first-ever financial penalty in this area, and the enforcement wave behind it, still depended on somebody noticing and saying something.
What happens when regulators monitor advertising with AI?
Enforcement today is rationed by detection cost. Finding non-compliant advertising across social, search, influencer and direct channels is expensive human work, and that expense, not tolerance, is why most breaches are never actioned.
Automated monitoring removes that constraint, and Australian regulators are already building toward it.
The TGA released its Compliance Principles for 2026 and 2027 on 22 January 2026, replacing its previous priority list with five core principles: Safeguarding therapeutic goods, Educate to empower, Protect those most at risk, Leverage digital capability, and Strengthen enforcement. Priority focus areas are now reviewed quarterly through intelligence-led, risk-based assessment. The framework is described as built for a rapid response to high-impact issues, including those arising from misinformation, digital platforms and online purchases.
The TGA also requested removal of more than 13,700 unlawful advertisements from digital platforms in FY 2024-25. That is not a volume a manual review team produces.
The ACCC's 2026-27 compliance and enforcement priorities, announced on 19 February 2026, include manipulative and false practices in digital markets.
The prediction: within the next six to twelve months, at least one Australian regulator moves from periodic sampling to continuous automated monitoring of digital and social advertising.
That's going to shake up a lot of brands that feel they can slip under the radar, especially on Instagram and other social media platforms.
Who this actually affects
Not primarily the largest advertisers. They are already visible, which is precisely why they get sampled, and they carry in-house legal teams.
It affects two groups.
- Small and mid-size brands with no in-house legal or compliance function. Their protection to date has been obscurity. Manual enforcement finds large advertisers because large advertisers are easy to find. Automated enforcement removes that asymmetry, because scanning the tenth-largest advertiser in a category costs the same as scanning the largest.
- Large brands running macro and micro influencer programmes. Legal capability does not help when content goes live at the speed of social. A review process measured in days cannot govern publishing measured in minutes.
Concentration of risk follows the sectors where the regulator is statutory rather than self-regulatory. Ad Standards research published on 15 July 2026 found community concern about influencer marketing is highest in health and medical products or advice, followed by financial products and services, gambling, and beauty and cosmetic services. Those categories map to the TGA, ASIC and ACMA, where the outcome of a finding is a civil penalty rather than a takedown. They are also heavy users of influencer marketing.
The AI disclosure problem arriving alongside it
Two further Ad Standards findings compound the picture.
Research published on 6 May 2026 found 72% of Australians are concerned about the use of AI-generated content in advertising, particularly its potential to mislead. One in five say they are not at all confident they could identify AI-generated content in an ad. 64% believe disclosure of AI-generated content should always be required, and Ad Standards has confirmed there are currently no specific rules requiring it.
Research published on 15 July 2026 found 84% believe it is important that influencers disclose when content is advertising, a requirement the AANA Code of Ethics already imposes through the rule that advertising must be clearly distinguishable as such.
So detection is degrading on the public side at the same moment it is expanding on the regulator side. Both trends push in the same direction. The check has to happen before publication, because nothing downstream is reliable any more.
What scales, and what does not
The only defence that matches always-on monitoring is screening every asset before it reaches final human review, or before it goes to market where full review is not affordable at all.
That screening has to be fast enough and cheap enough that it actually happens on everything, not just on the assets someone remembers to flag.
This is the position Hiaitus works from.
Human Intelligence multiplied by AI. Purpose-built systems at the point where judgement happens, with a person making the final call.
A general-purpose AI assistant asked to check an ad is a pattern matcher with no grounding in ABAC Part 3, AANA Wagering Code section 2.5 or Therapeutic Goods Act section 42DLB, and no way to tell you which rule it applied.
Checked It is the applied case.
- Checked It pre-screens marketing creative against Australian regulatory frameworks across alcohol, gambling, supplements, insurance, news and food and beverage.
- Checked It returns findings against specific code sections rather than general impressions, and it retains the audit trail.
- Checked It is not legal advice and not a substitute for human review.
- Checked It is the layer that makes sure the last check you control actually runs, on every asset, at the speed the work moves.
Get a Free Trial at www.checkedit.ai
Frequently asked questions
How many influencer posts did the ACCC review in its sweep?
118 posts, across Instagram, TikTok, Snapchat, YouTube, Facebook and Twitch, in its January 2023 internet sweep. 81% raised concerns under the Australian Consumer Law.
Has the ACCC fined a brand for undisclosed influencer content?
Yes. On 24 March 2026 the ACCC issued two infringement notices over allegedly misleading influencer reviews on Instagram, with $39,600 in penalties paid. It was the first financial penalty of its kind in Australia. The ACCC notes that payment of an infringement notice penalty is not an admission of a contravention.
Do influencers have to disclose paid partnerships in Australia?
Yes. The Australian Consumer Law prohibits misleading or deceptive conduct, and the AANA Code of Ethics requires advertising to be clearly distinguishable as advertising. Gifted product counts as payment.
Are Australian regulators using AI to monitor advertising?
Regulators are building digital monitoring capability. The TGA lists "Leverage digital capability" among its five compliance principles for 2026 and 2027, reviews priority focus areas quarterly through intelligence-led assessment, and requested removal of more than 13,700 unlawful advertisements from digital platforms in FY 2024-25. No Australian advertising regulator has publicly confirmed AI-assisted advertising surveillance at scale.
Which sectors face the highest advertising compliance risk in Australia?
Sectors with a statutory regulator rather than only a self-regulatory code: health and supplements under the TGA, financial services under ASIC, and gambling under ACMA. These are also the sectors where Australian community concern about influencer marketing is highest.
Sources
- ACCC internet sweep of influencer content, January 2023, and associated ACCC reporting
- ACCC media release, 24 March 2026
- Therapeutic Goods Administration, Compliance Principles 2026 and 2027, released 22 January 2026
- Therapeutic Goods Administration media release on updated social media advertising guidance, November 2025
- ACCC 2026-27 Compliance and Enforcement Priorities, announced 19 February 2026
- Roy Morgan for Ad Standards, Community Sentiment Tracker, published 6 May 2026, fieldwork 19 February to 2 March 2026, n=1,000
- Roy Morgan for Ad Standards, Community Sentiment Tracker, published 15 July 2026, fieldwork 18 May to 1 June 2026, n=1,006
- AANA Code of Ethics
This article is a general overview for marketing leads in regulated sectors. It is not legal advice.