Monthly Digital Insights – September 2026

AI is becoming an active participant in marketing, acquisition and measurement, while developments in attribution and privacy continue to challenge existing analytics frameworks. This month, we look at AI-driven advertising and activation, Google's /goto redirect and its impact on attribution, and why strong privacy and data governance matter more than ever.

From implementation to retargeting, AI is everywhere

AI is rapidly moving from an assistant behind the scenes to an active participant in acquisition, activation and measurement. OpenAI introduced ChatGPT Ads into the European market, complete with measurement capabilities including OpenAI Pixel and Conversions API. At the same time, vendors such as Tealium and Hightouch are building agentic capabilities that allow AI agents to interact directly with Customer Data Platforms. To top it off, the European Commission has officially designated ChatGPT as a Very Large Online Search Engine (VLOSE) under the Digital Services Act.

AI is well beyond content generation; we are shifting from "AI helps marketers" to "AI participates in marketing". It is becoming a discovery platform, an advertising channel and, increasingly, a participant in customer journeys. For analysts, this raises new questions around attribution, audience activation, consent and measurement.

Why this matters

Organisations should start evaluating how AI-driven traffic, agent interactions and AI-powered activation fit into their measurement frameworks. Just as social media became a new traffic source a decade ago, AI platforms are now becoming an important category in attribution models.

Monitoring Google's /goto redirect for attribution

Google recently started routing search result clicks through a new google.com/goto redirect endpoint, triggering concerns across the analytics community about potential attribution issues. However, Attributer.io conducted some testing, and concluded that referrer information remains intact and traffic continues to be attributed correctly as Organic Search.

When Google changes click-routing behaviour, analysts should pay attention. A change in referrer handling can have major consequences for attribution models, channel reporting and budget allocation decisions.

Why this matters

The most important lesson is not whether the redirect itself is problematic, but that organisations should continuously validate attribution assumptions. Changes in browser behaviour, privacy features and platform infrastructure can have unexpected side effects.

Monitoring Organic versus Direct traffic remains good practice, but at this stage we found no evidence that the /goto rollout requires remediation.

Context matters more than ever for privacy and governance

Two privacy-related stories caught our attention this month. First, the Dutch Data Protection Authority (AP) fined Uber nearly €825 million for making impactful decisions about drivers through automated systems without meaningful human intervention. Second, the discussion around GA4Spy sparked debate about whether Google Tag Manager implementations expose too much information publicly.

Although these stories appear unrelated, they share the same message: regulators care about how organisations use data, while tools such as GA4Spy remind us how visible those data practices are becoming.

The Uber decision reinforces that organisations cannot blindly rely on automated decision-making where individuals are significantly affected and human oversight remains a core GDPR requirement. GA4Spy is a reminder that, even though marketing technology stacks have always been partially visible through browser DevTools and consent banners, it just became easier to inspect and flag improper implementation.

Why this matters

The GA4Spy discussion feels somewhat overblown. Client-side tags have never been private, and most marketing partners can already be identified through consent management platforms and browser developer tools.

The real question is not whether your Tag Management System can be inspected, but whether your governance, consent implementation and tagging strategy are robust.

At the same time, the Uber ruling is an important reminder that regulators continue to focus on the human consequences of data and AI systems. As organisations deploy more AI-driven workflows, governance should evolve alongside technical innovation.

Key takeaway

This month's stories perfectly illustrate a point we recently discussed at Data Expo: context is becoming the most valuable asset in modern analytics.

The challenge isn't collecting more and more data. It's about governing increasingly complex ecosystems as AI platforms become traffic sources, AI agents become users and regulators become more interested in how automated systems influence people.

Context is everything, especially now that platforms are implementing MCP and agentic capabilities to facilitate insights and decisioning.

Success increasingly depends on understanding context, governing data responsibly and designing measurement frameworks that can keep pace with AI-driven interactions.

Receive data insights, practical cases and a look behind the scenes once a month?

Sign up to our mailing list and stay 'up to data':