The news: Meta has indefinitely shelved a proposal that would have allowed Instagram and Facebook to scrape websites operated by Meta Business account holders to train its AI models.

The pause comes after Meta delayed the update from the end of October to mid-November to “explain to our partners and stakeholders what is changing” based on a slew of complaints from marketers and publishers as well as legal action in Norway, Brazil, and the UK. In the US, Senator Adam Schiff (D-CA) called the proposal “a gross abuse of Meta’s power.”

Why it matters: The delay doesn’t resolve underlying concerns.

The proposal raises broader questions about how much control advertisers retain over their own content and business data when using major platforms, and what rights those platforms have to access and use it.

The incident also points to a broader tension between emerging AI-driven advertising products and brands’ control over their own data and intellectual property.

Marketers have traditionally shared first-party data with major platforms for targeting and optimization. But as platforms gain more control over advertising decisions, brands are increasingly sharing valuable signals like transactions and offline outcomes with systems they do not fully control.

That suggests that, more broadly, transparency and control are becoming influential factors in where marketers allocate budgets—potentially making platforms’ data-use policies a greater consideration when evaluating advertising partners. Nearly 64% of US brand and agency leaders said they had shifted mobile ad budgets from walled gardens to independent partners in the 12 months preceding July, compared with just 1.2% who moved budgets in the opposite direction, per Datum Intelligence and EMARKETER.