This time it feels different. It’s not because another regulation is coming or because platforms are rewriting the technology map. This is because the environment surrounding data itself is changing.

At least, that’s the direction the events of the past week suggest.

TikTok’s U.S. privacy policy was updated last week to clarify that the app may collect precise location data if users give their permission. That in itself is not noteworthy. After all, detailed behavioral signals are the currency of the modern platform economy. But context is important. TikTok in the US now operates under a new ownership structure that includes Oracle, whose co-founder Larry Ellison has talked about AI-powered surveillance as a tool to keep the public on their “best behavior.” Against such rhetoric, everyday data practices take on a different weight.

At the same time, U.S. Immigration and Customs Enforcement is seeking information from advertising executives about how “big data” and “ad tech” can be used to aid investigations. In other words, tools built to predict who will buy something can be used to research buyers.

Taken together, these developments will reshape the role of advertising in the data ecosystem. The industry has helped normalize the continuous collection of behavioral signals at scale. For many years, these signals were treated as innocuous input for correlation and measurement. But they have been aggregated and modeled, forming a detailed and permanent record of everyday life, an asset whose value now goes beyond marketing.

And that’s before AI comes into the equation. Systems can now infer attributes, risks, and future behavior with statistical confidence, even when those attributes are not explicitly shared. Unintended consequences are a concern. The existence of these predictive capabilities rapidly expands opportunities for abuses such as fraud, discrimination, manipulation, and state surveillance.

“What we’re talking about is the unintended social consequences of algorithms knowing more about us than we know ourselves,” said Jamie Barnard, CEO of Compliant, a startup that helps companies comply with privacy laws. “Look at how the lines between consumerism and state surveillance are blurring in the most alarming and frightening ways in China. This is a warning of this endpoint.”

For marketers who create dilemmas that can’t be easily resolved with compliance checklists. No brand team is trying to enable surveillance. But after years of pushing for clearer targeting, richer measurement, and persistent identity, we have been able to create the data environment that AI now feeds on. The same interoperability that streamlines advertising also allows companies to port data in ways they don’t have full control over.

Ad tech companies face similar constraints. Its value lies in sending a signal at scale. Advertisers pay for that connection. This is also what makes the data powerful in other contexts. The risk Barnard frames is not just that companies hold data. That means actors with very different incentives will have direct or indirect access to aggregated datasets.

If that sounds alarming, take a look at how location data has surfaced in criminal investigations, how social data has been used to assess creditworthiness and risk of a case, or how predictive analytics has underpinned everything from fraud detection to border enforcement.

“We are entering a new paradigm where advertising infrastructure intersects more with society and democracy,” said Alex Tait, founder of digital consultancy Entropy and former media director at Unilever. “The issue isn’t awareness; it’s leadership. In markets like the UK, where new electoral and online safety laws are on the horizon, the industry has a real opportunity to share expertise and help shape how these systems are managed. That opportunity is also emerging internationally, as governments increasingly treat advertising infrastructure as democratically important.”

Where this will ultimately lead remains to be seen. What is also clear is that the industry is moving to a point where second-order effects of data practices can no longer be treated as distant hypotheses. Systems built for performance are now part of a broader information ecosystem shaped by AI, geopolitics, and state power.

“The test for the next generation of industry leaders is whether they step up and apply their expertise to building solutions, or whether they continue to treat some of these problems as thorny issues,” Tate said.

The real differentiator in ad tech today is not technology.

There is much to acknowledge in this industry, but integrity has never been its calling card. In moments like now, that is painfully obvious. That’s why I’m always grateful for reminders, big or small, of the relationships that really hold this business together.

That thought arose in a conversation between industry expert Gina Cavallo, now chief revenue officer at French measurement company XPLN.AI, and Michael Lampert, senior director of global generative AI implementation and consumer data strategy at Mondelez International. Rather than trying to wallow in sentimentality, such moments land as necessary palate cleansers in the nonstop parade of heartthrob, anxiety masquerading as expertise, and, here, pure selfishness that often masquerades as thought leadership.

This made what happened next all the more remarkable. The discussion was supposed to be about the product: attention metrics, creative effectiveness, CTV, but the subtext kept stealing the show. Before the XPLN.AI methodology aired, Mr. Lampert provided something close to a character reference, positioning Mr. Cavallo as a trusted advisor in the market who would serve as a secondary to the bright features on that week’s slide deck. He spoke out loud a truth that marketers rarely articulate. “There are only five or 10 people that I really trust enough, especially in a space like this, that I’d even want them in the building. That’s because trust is important.”

Emerging fields such as attention and AI-powered measurement lack decades of evidence and trust is the transit point.

That perspective reconstructs Cavallo’s movements. With stops spanning Experian (via Audigent), MiQ, Playground XYZ, and LiveRamp, she had options. Instead, she chose to build a U.S. business for a European attention measurement company, taking on the difficult challenge of growing a startup rather than leveraging the scale of a platform. Its mission is structural. That means turning a few seconds of attention into something marketers can use to adjust their ongoing media, linking creative effectiveness with media efficiency, at a time when brands are under new pressure to prove that their investments are delivering real results, not just results.

Still, if this conversation is any guide, XPLN.AI’s early advantage in the U.S. may be more relational than technological. In her words, “relationships open doors,” but the product has to justify staying in the room. In a noisy hype cycle, a trust-first, technology-second order looks like a more durable sign.

Numbers you need to know

71%: Percentage of German children whose most popular VoD platform is Netflix

21%: Reels as a percentage of total Instagram ad impressions in Q2 2025

20%: Increase in percentage of time spent in meta threads in Q4 2025

50%: Capacity growth rate for OpenAI models powering Microsoft Copilot workloads

What we covered

TikTok’s US deal still leaves unanswered questions

TikTok has finally come to light, or so everyone thought when the deal with China was confirmed. But as with any announcement, questions remain about how this will play out.

‘There seems to be a shift in mindset’: Advertisers keep ad spend flexible as uncertainty continues

No one would disagree that last year was a turbulent year. And marketers were expecting a more predictable kind of disruption in 2026. Spending plans are wrapped in cautious optimism, but they are driven less by belief in a stable economy than by faith in one’s ability to weather an economy that is not.

Advertising looks like the next risky move as ChatGPT growth slows

Starting next month, OpenAI will begin testing ads on ChatGPT in the US, but it appears not all users are happy with the idea. Some users have openly claimed that they will leave the platform if they see an ad.

Inside NBCU’s $3 million Peacock Super Bowl pitch

Digiday spoke to media buyers to share all the details about what the NBCU platform is looking for from advertisers for the 30-second ads it will run during Peacock’s Super Bowl broadcast.

what we are reading

Nvidia, Microsoft, Amazon in talks to invest up to $60 billion in OpenAI

According to The Information, OpenAI is seeking to raise up to $100 billion to fund its AI, and tech giants Nvidia, Microsoft, and Amazon are in talks to invest up to $30 billion, $10 billion, and more than $20 billion, respectively.

Google’s scupper service compares YouTube viewing to TV and streaming viewing

According to the Financial Times, Google has sent a “cease and desist” letter to Kantar Media over its new measurement service that would prevent it from accessing data that would help it compare YouTube channel viewership to other TV channels and streamers and attribute views to specific content creators.

Welcome to desocialized media

According to New York magazine, the platform once built to connect people around the world is now focused on leveraging its algorithms to prioritize endless and isolated video consumption.

TikTok creator Kirby Rame signs $975 million deal with his brand

Kirby Lame, the most followed TikToker, has sold his company’s stock to Hong Kong-based financial firm Rich Sparkle Holdings in an all-stock deal worth $975 million, according to The Hollywood Reporter.