cd /news/ai-safety/metas-customer-disservice-bots-payin… · home topics ai-safety article
[ARTICLE · art-19394] src=adexchanger.com pub= topic=ai-safety verified=true sentiment=↓ negative

Meta’s Customer Disservice Bots; Paying For Performance

Meta's AI customer support chatbot, rolled out across the ads platform in March, was exploited by hackers who convinced it to grant them admin access to high-profile Instagram accounts, including those of Barack Obama and Sephora. The hackers simply told the chatbot they were the true account owners, prompting it to link new email addresses as administrators without human verification. Meta has since patched the exploit, but the breach highlights ongoing failures in the company's automated customer service system.

read4 min publishedJun 2, 2026

Would You Like Help?

One way Meta has pushed adoption of its AI products is by touting their usefulness for business accounts, including features like video ad creation tools and agentic customer support.

So it’s probably not great that the larger Meta AI support chatbot, which rolled out across the entire ads platform in March, launched with a sizable security hole.

As 404 Media reports, a group of hackers claimed they’d hijacked a number of high-profile Instagram accounts, including ones for Sephora, Barack Obama and the Chief Master Sergeant of the Space Force.

Apparently, the hackers simply convinced Meta’s AI support that they were the true owners of the Instagram accounts in question. The chatbot of its own authority would then link new email addresses as admins for the accounts.

Although Meta did not respond to 404 Media’s inquiries, the company has seemingly patched the exploit. In the meantime, though, several of the hacks (that we know about, at least) made national news – or, in Sephora’s case, got noticed by customers posting about the takeover on Reddit.

Meta’s customer service has been so bad for years that illegal customer service schemes have cropped up just so folks can have their normal account issues seen to.

So don’t expect much here. Advertiser complaints can go to the chatbot.

Luxury Performance

The luxury marketing agency Interluxe Group has “dabbled” in performance marketing, Founder and CEO Nick Van Sicklen tells ModernRetail.

Which will change in a big way now that Interluxe has acquired adMixt, a data-driven performance marketing agency.

Luxury brands (Interluxe’s clientele includes the Four Seasons, Rolls-Royce and Ferragamo) have mostly eschewed standard performance marketing as being cheap and not, well, luxe enough. But now, even the most upper-funnel branding companies must bring a data-driven, results-oriented perspective.

Alex Greifeld, an ecommerce growth consultant, notes that when most brands add a channel, such as embracing Amazon or user-generated content on Instagram, they see immediate strong results. (One reason is because they’re picking up the easy, preexisting demand, not true new customers, but anyway.)

But for companies that sell at super-high premiums – like multimillion-dollar homes or watches worth tens of thousands of dollars – it’s not always easy to find the very particular audiences they need using the tactics of established luxury marketing agencies, which prioritize experiential branding, high-production video and direct media deals.

AdMixt will no doubt mixt in some new data-driven ads for Interluxe.

Disagreement On Principal

Advertiser trust in agencies hasn’t improved much in the past decade, according to a survey released Monday by the Association of National Advertisers and research firm K2.

Forty-three percent of ANA members are concerned about a lack of transparency from their agency partners, the survey found. In a previous ANA and K2 survey from 2016, that number was 46%.

The ANA was “absolutely hoping for more progress,” ANA group EVP Bill Duggan tells Ad Age. But agencies collecting cash rebates from ad sales remains a major unchecked source of mistrust, according to ANA members. And principal media deals, in which agencies resell ad inventory to buyers and keep the margin for themselves, are another growing concern.

However, agencies aren’t shying away from principal media, and it’s a booming business. Ad Age notes that WPP made $713 million from principal media in 2024, according to filings by former GroupM exec Richard Foster in his lawsuit against WPP.

Meanwhile, brands aren’t exactly getting more transparency as agencies increase their commitments to walled garden platforms, argues Nick Manning, who worked on this year’s ANA/K2 survey. And with advertising growth concentrated in areas controlled by the big platforms, including social media, CTV and retail media, don’t expect the problem to fix itself anytime soon.

But Wait! There’s More!

What Google’s new AI actually debunks and what it doesn’t. [Search Engine Journal] Meanwhile, web visits and app installs of DuckDuckGo’s AI-free search page have increased exponentially since Google announced plans to overhaul its service. [TechCrunch]

Anthropic confidentially files for its IPO. [Reuters] YouTuber Dhar Mann is going to Tribeca X to prove that CTV can do TV’s job. [Digiday]

A new technique might allow websites to spy on user activity by measuring interactions in their solid-state drives. [Ars Technica]

Minecraft launches its first-ever affiliate program, powered by impact.com. [release]

You’re Hired!

Expedia Group appoints Bill Watkins to lead global advertising. [release]

── more in #ai-safety 4 stories · sorted by recency
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
Live at https://your-agent.zahid.host
Get free account → Pricing
from €0/mo · no card required
LIVE [news/metas-customer-disse…] indexed:0 read:4min 2026-06-02 ·