Opinion | Big Tech Fooled America Once. The Second Time’s Not Going So Well.
By Gammatek ISPL Last updated: September 2026 | 13 min read
Why This Matters
Fifteen years ago, Silicon Valley made America a promise: give us your data, your attention, and your trust, and we'll connect you to the world, organize its information, and make your life easier. A critical mass of the public said yes. Today, the same industry — often the very same companies — is making a new promise about AI: give us your data, your jobs' workflows, and your trust again, and we'll make you more capable, more informed, more free. This time, a much larger share of the public is saying "we've heard this before." Understanding why that skepticism formed, and whether it's justified, matters because it will shape how fast AI actually gets adopted, how it gets regulated, and whether the companies building it can recover the credibility they spent the last decade losing.
The First Pitch, and What It Actually Delivered
The social media era's founding promise was genuinely appealing: free tools that connected people across distance, gave small businesses a marketing platform they could never have afforded before, and gave ordinary people a voice that used to belong only to publishers and broadcasters. For years, that promise mostly held up in the public imagination.
Then the costs became visible. Data harvesting practices exposed by journalists and whistleblowers. Business models built on maximizing attention, not wellbeing. Content moderation decisions that satisfied no one and appeared inconsistent or self-serving. Congressional hearings that revealed executives who seemed not to fully understand their own products' effects. By the time the dust settled, a broad, bipartisan skepticism of "Big Tech" had set in — not a fringe position, but something close to consensus that these companies had not been straight with the public about what they were building or why.
The Second Pitch
Now the same core group of companies — plus a handful of well-funded newcomers — are asking for trust again, this time for AI. The pitch has familiar shape: this technology will make you more productive, more creative, more informed, and it's inevitable, so better to adopt early than get left behind.
What's different this time is the starting position. In the 2010s, the public extended trust by default and had to be shown reasons to withdraw it. In 2026, a meaningful share of the public starts from skepticism and has to be persuaded to extend trust at all. That's a much harder sales environment, and it shows in the data: surveys have consistently found Americans more anxious than optimistic about AI's effects on jobs, privacy, and misinformation, even as usage of AI tools themselves continues to climb. People are trying the products while doubting the companies selling them — a strange, unstable combination that doesn't typically last.
Where the Comparison Holds Up
A few parallels between the two eras are genuinely strong, not just rhetorical:
The "inevitability" framing is identical. Just as Facebook and Google once argued that network effects made their dominance natural and unstoppable, AI companies now argue that competitive and geopolitical pressure make rapid, under-regulated deployment necessary — that slowing down means falling behind, whether to a rival company or a rival nation.
The self-regulation pitch is identical. Social media companies spent years arguing they could police themselves better than regulators could, right up until scandals proved otherwise. AI companies are making structurally the same argument today — voluntary safety commitments, internal review boards, self-published safety research — before any binding external framework exists.
The scale-first, fix-it-later pattern is identical. Social platforms shipped features and dealt with harms after they surfaced at scale. Several major AI deployments have followed a similar pattern — releasing capable systems broadly, then patching guardrails in response to problems discovered by users in the wild rather than caught beforehand.
Where the Comparison Breaks Down
It would be dishonest to pretend these are identical situations, and a fair-minded reader deserves the differences too:
The public is not starting from zero this time. Precisely because of the social media experience, journalists, regulators, and ordinary users are scrutinizing AI companies' claims faster and more skeptically than they scrutinized the last generation. Congressional hearings on AI happened years before mass deployment, not years after — a genuinely different sequence than the social media story.
Some AI companies have made real structural changes in response to lessons from the social media era — more external safety research partnerships, more willingness to publish uncomfortable findings about their own systems, clearer usage policies. Whether these changes are sufficient is a fair debate, but treating every AI company as identical to social media's worst actors would flatten real distinctions worth preserving.
The business models aren't identical. Social media's core harms were tightly linked to an advertising-attention business model that rewarded engagement over wellbeing by design. Several major AI products monetize through direct subscriptions or enterprise licensing rather than attention-harvesting advertising — a structural difference that changes (though doesn't eliminate) the incentive problems critics point to.
[IMAGE — two-column comparison: "What's the Same" vs "What's Different" between the social media trust crisis and the current AI trust environment] Alt text: "Two-column comparison chart showing similarities and differences between the social media era trust crisis and the current AI industry trust challenges" Caption: "Some patterns are repeating. Others are genuinely new — a fair assessment has to hold both at once."
A Data Point Worth Sitting With
One useful signal: independent research (including Pew Research Center surveys tracked across 2023-2026) has repeatedly found that a majority of Americans favor more regulation of AI companies, while a much smaller share believes those companies will act responsibly on their own — a gap between "we don't trust you to self-police" and "we're using your product anyway" that didn't exist at nearly the same scale during social media's early growth years. That gap is the real story: skepticism and adoption are rising together, not trading off against each other the way conventional market logic would predict.
An Implementation Consideration for Anyone Evaluating AI Vendors
If you're a business leader deciding whether to adopt an AI vendor's tools, the social-media-era lesson worth applying directly is this: read the vendor's actual terms on data usage and retention before adoption, not after a problem surfaces. Social media's worst trust failures were often things that were technically disclosed in terms of service nobody read, not hidden entirely. The AI era offers a chance to actually do that due diligence up front, since the industry-wide skepticism has already put these questions on the table earlier than it did last time.
What Happens If Trust Doesn't Recover
If the current gap between adoption and trust persists, a few plausible outcomes: faster and more assertive regulation than the industry would prefer, since a public that doesn't trust self-policing tends to support external rules; slower enterprise adoption in trust-sensitive sectors (healthcare, finance, anything regulated) than the technology's raw capability would otherwise predict; and a market opening for vendors who differentiate specifically on transparency and auditability rather than raw capability — a lane social media companies never seriously competed in, because attention-based business models didn't reward it.
That last point is where this connects to something concrete: for any organization built on trust-sensitive data — plant safety records, compliance documentation, employee information — vendor transparency isn't an abstract virtue, it's the actual criterion procurement teams are increasingly asked to evaluate before signing a contract.
The Fair Conclusion
Big Tech's first pitch worked because the public hadn't yet learned what large-scale data collection and attention-optimized products actually cost. The second pitch is landing in a different environment — not because the technology is necessarily less trustworthy, but because the audience has already paid the tuition once and isn't inclined to skip the homework this time. Whether that skepticism is fully justified or somewhat overcorrected is a genuinely open question — reasonable people land in different places on it — but the skepticism itself is not irrational. It's earned. https://www.gammateksolutions.com/post/2026-price-comparison-of-hci-hyper-converged-infrastructure-solutions https://www.gammateksolutions.com/post/openai-playground-explained-how-it-works https://www.gammateksolutions.com/post/fortinet-cyber-security-pricing-2026-firewall-cost-guide
Comments