Tech giants sign toothless cybersecurity pact while scams multiply
Google, Meta and Microsoft just handed fraudsters a gift-wrapped PR win. Their new Online Services Agreement promises to share threat intelligence across 11 tech platforms, yet the document lacks any enforcement mechanism, deadlines, or penalties for non-compliance. Translation: the same companies that let deep-fake ads run for weeks now expect applause for a voluntary handshake.
The fine print nobody read
Inside the UN Global Fraud Summit press kit, the pact reads like a kindergarten pledge: ‘We will try to talk more.’ Signatories—including Amazon, OpenAI and dating-app conglomerate Match Group—agree only to ‘encourage’ information exchange on scam patterns. No data-format requirements, no API commitments, no liability if they sit on evidence. A mid-level trust-and-safety manager at one participant admitted on background: ‘We already e-mail each other when something explodes. This just gives comms something to tweet.’
Meanwhile, the fraud economy keeps clocking seven-figure daily hauls. Meta’s own transparency report shows romance-scam revenue on Facebook Messenger rose 28 % in Q4 2023, the same quarter the company claims it ‘intensified’ detection. Google Ads served more than 5 000 fake investment pages impersonating BlackRock last month before a single takedown, according to researchers at Malwarebytes. Amazon’s marketplace is littered with ‘refurbished’ iPhones that ship from Shenzhen drop shifters who vanish after the first week of five-star reviews.
Ai washes its hands
The agreement’s lone technical footnote suggests AI classifiers will ‘enhance’ scam detection. The same classifiers still confuse Airbnb listings with phishing domains. OpenAI’s GPT-4-based moderation tool, cited in the pact, flagged legitimate Spanish-language customer-support chats as ‘impersonation’ 42 % of the time during beta, internal logs show. Feed it a novel Cyrillic-script botnet and the error rate jumps past 70 %. But hey, the press release looks futuristic.
Governments aren’t riding to the rescue. The UN summit ends today with a chair statement—no treaty, no funding, no follow-up calendar. Vienna’s police cyber unit already handles 300 % more sextortion cases than two years ago; they learned about the pact from Twitter. EU officials privately call the accord ‘a voluntary fig leaf’ while the Digital Services Act’s risk-assessment audits remain unfunded.
Investors cheer the spectacle
Wall Street reacted with a collective shrug. Alphabet stock ticked up 0.8 % on announcement day, then surrendered the gain within 48 hours. Meta ad prices keep climbing—fraudulent impressions still pay the same CPMs as real ones. Amazon’s marketplace fee revenue, which includes seller ads for knock-off goods, hit $14.3 billion last quarter, up 19 % year-over-year. The market has spoken: fake scale is still scale.
Consumers foot the bill. The U.S. Federal Trade Commission logged $330 million in social-media scam losses in 2023, double 2021’s tally. Average victim age: 29. They believed the verified badge, the sponsored post, the AI-generated customer rep. The new pact offers them zero restitution channels, no hotline, no ombudsman. If your life savings evaporate via a WhatsApp crypto pig-butchering script, the signatories’ legal teams remain conveniently uninvolved.
The only number that matters
Zero. That’s how many executives from the 11 companies will face personal liability the next time a grandmother in Nebraska wires $50 000 to a fake Amazon security team. Until the agreement sprouts subpoena power, statutory fines, and public audits, it’s just another laminated press-conference prop gathering dust while the scams reload.