AI-Powered Investment Scams Cost Australians $837.7M in 2025 and Regulators Are Losing the Takedown Race

Australia's corporate regulator removed 11,964 investment scam and phishing websites in 2025, a 90% increase on the prior year, yet scam losses still reached $837.7 million across the twelve months, according to data released today by the Australian Securities and Investments Commission (ASIC) and the National Anti-Scam Centre. The figures expose a compounding dynamic: takedown capacity is scaling, but so is the threat.

Total scam losses across all categories reached $2.18 billion in Australia in 2025. Investment scams accounted for the single largest component, driven by an accelerating use of AI-generated content to manufacture convincing fake endorsements, professional-grade video advertising, and geo-targeted social media campaigns designed to evade

platform moderation.

How AI Has Changed the Scam Threat Profile

The operational shift ASIC is describing is not incremental. Scammers are now deploying large language models and synthetic media tools to produce content that is materially harder to distinguish from legitimate financial advertising, polished video, celebrity deepfakes, and algorithmically targeted ad placement that exploits consumer interest in AI-driven investment returns.

ASIC Commissioner Alan Kirkland identified a specific evasion technique gaining traction: cloaking, where scam advertisements display different content depending on the user's device, location, or browser profile. The method allows fraudulent ads to pass platform review while serving harmful content to targeted consumers. "Scammers are using artificial intelligence to make fake investment ads look more polished, more convincing and harder to spot," Kirkland said.

The tactical playbook is consistent across cases ASIC has identified: claims of AI-powered trading bots generating passive income, fake regulatory certifications listing Australian Company Numbers (ACNs) belonging to unrelated entities, and fabricated news pages carrying AI-generated endorsements from prominent Australians.

The Takedown Numbers in Context

Since ASIC launched its scam takedown service in 2023, more than 25,000 investment scam and phishing websites have been removed. In 2025 alone, the regulator averaged 32 takedowns per day, or roughly 230 per week. ASIC also removed more than 1,100 social media investment scam advertisements during the year.

Those are meaningful operational numbers. But set against $837.7 million in investment scam losses in a single year, the cost-per-takedown arithmetic is sobering. Each removed website represents a fraction of a cent of loss prevention if the underlying fraud infrastructure simply migrates to new domains, which, given the low cost of domain registration and AI content generation, it readily can.

ASIC acknowledges the arms-race dynamic implicitly: the regulator notes that some scam ads appear in social media feeds for only short periods before being taken down and reposted under a different appearance, requiring 24/7 monitoring through a third-party cybercrime detection partner.

What the Regulatory Response Looks Like

ASIC's operational model relies on a third-party specialising in cybercrime detection and disruption, which proactively identifies both scam websites and social media advertisements directing consumers to fraudulent platforms. Once malicious activity is confirmed, ASIC coordinates takedown requests across hosting providers, registrars, and social media platforms.

The regulator is also engaged in international cooperation with overseas counterparts, a necessary posture given that scam proceeds frequently flow to overseas criminal networks, as Commissioner Kirkland noted. The cross-border dimension of investment fraud significantly complicates enforcement: Australian consumer losses, foreign perpetrators, and infrastructure distributed across multiple jurisdictions.

ASIC's consumer-facing guidance centres on AFS licence verification, investors are advised to confirm that any entity soliciting funds holds a valid Australian Financial Services licence before engaging. The regulator's Moneysmart investor alert list and "Check before you invest" resource are the primary public-facing tools.

Why the $837.7M Figure Understates the True Cost

Reported scam losses are a floor, not a ceiling. Consumer research consistently shows that investment scam victims under-report due to embarrassment, lack of awareness of reporting channels, or the belief that reporting will not result in fund recovery. The National Anti-Scam Centre's Targeting Scams Report, which supplies ASIC's headline figures, draws on voluntary reports to Scamwatch and does not capture unreported losses.

Independent estimates from consumer advocacy groups have historically placed actual scam losses at two to three times reported figures. If that ratio holds, Australia's true investment scam burden in 2025 may be closer to $1.7–2.5 billion, a figure that would represent a systemic financial stability concern, not merely a consumer protection issue.

For institutional investors and wealth managers, the reputational contamination risk is also relevant: AI-generated scam content frequently mimics the branding, language, and investment propositions of legitimate financial products, eroding consumer trust in the sector broadly.

Why This Matters to FinanceX Readers

For compliance officers, wealth managers, and fintech operators, ASIC's data is a directional signal with three practical implications.

First, AI-generated fraud is now a baseline compliance risk, not an edge case. Any financial institution with a consumer-facing digital presence needs to actively monitor for brand impersonation and fake endorsement content, not wait for ASIC to flag it.

Second, the regulatory scrutiny on social media platforms is intensifying. ASIC's 1,100+ social media ad takedowns in 2025 indicates the regulator is actively engaging with platforms on content moderation standards. Financial services advertisers operating on the same platforms face heightened due diligence expectations on ad content and targeting parameters.

Third, the gap between takedown volume and loss figures suggests that reactive removal alone is insufficient. Firms investing in proactive fraud detection infrastructure—both to protect their own brands and their customers, are likely to face lower regulatory and reputational risk as ASIC's enforcement posture continues to harden.

The 90% increase in takedowns in a single year is not a success story. It is a measure of how rapidly the problem scaled.

By Koen Vanderhoydonk - FinanceX Magazine