This article is for informational purposes only and does not constitute tax, legal, or accounting advice. Consult a qualified tax professional before making compliance decisions.

Policy

Pennsylvania's Digital Ad Tax Clears the House 139-63: The Gross-Receipts Template Just Won Its First Floor Vote

Beardsley Rumble|2026-06-12|5 min read

Eight days ago I wrote that Pennsylvania had picked the Maryland model over the Illinois one: a gross-receipts digital advertising tax defined by category rather than a bespoke tax defined by "machine learning algorithms." I flagged one signal to watch above all others — whether HB 1678 would clear the Pennsylvania House floor or stall in budget negotiations. On June 10, 2026, it cleared, 139-63. That vote is the development, and the margin is the story.

This is not a re-run of my June 4 analysis of the committee vote. A party-line committee advance tells you what a majority caucus wants. A 139-63 floor vote with 39 Republicans crossing over to join the Democratic majority tells you something structurally different: the gross-receipts template can assemble a bipartisan supermajority. For anyone building AI agents that buy, place, or serve advertising, that durability signal matters more than the rate.

What Actually Happened

HB 1678, sponsored by Rep. Elizabeth Fiedler (D-Philadelphia), passed the Pennsylvania House on June 10, 2026 by a 139-63 vote. Thirty-nine Republicans joined nearly all Democrats; only two Democrats voted against. The bill extends Pennsylvania's existing 5% gross receipts tax to digital advertising services — banner, search, interstitial, and similar — provided by large platforms with in-state receipts. The Independent Fiscal Office projects between $329 million and $500 million in revenue for fiscal 2026-2027, earmarked in the House framing toward senior property tax relief. Local news media and broadcast outlets are exempt.

The bill now moves to the Pennsylvania Senate, with the state's budget deadline at the end of June. It still needs Senate passage and Governor Josh Shapiro's signature before it is an enforceable obligation. So my comfort framing has not changed: on whether HB 1678 becomes law as written, I remain at Reasonable Basis at most. A House passage is real progress, but Pennsylvania budget bills are amended, traded, and stripped in the Senate every cycle, and digital ad taxes have a particular history of dying late.

Why the Margin Changes the Analysis

What I now hold with Substantial Authority is sharper than it was a week ago. On June 4 I argued that the gross-receipts model — Maryland's design, now Pennsylvania's — was the template legislators reach for when they want revenue that can survive a courtroom. The 139-63 vote adds a second dimension to that thesis: the gross-receipts model is also the one that can survive a floor.

Consider the contrast with Illinois. Illinois enacted its 10% Targeted Advertising Services Tax (SB 3019) as a late amendment folded into a budget bill — a mechanism that never required members to cast a clean, standalone vote on taxing automated advertising by its "machine learning" delivery method. Pennsylvania did the opposite. HB 1678 got an up-or-down floor vote on a category-based gross-receipts structure, and it drew 39 Republican votes. A tax base that 39 members of the opposing party will publicly endorse is a tax base that travels. A tax base defined by "machine learning algorithms" — which reaches into the architecture of the systems it taxes — is a far harder vote to whip across the aisle.

That is the contagion mechanism agent operators should internalize. States copy what passes. Two states have now moved on digital advertising in a single cycle, and the one that produced a bipartisan floor supermajority used the model that does not key its base to how an algorithm works.

What This Means for AI Agents

The agent economy is absorbing the advertising function — agents that run paid acquisition, bid in real-time auctions, generate creative, and allocate budget across channels. Under the gross-receipts template, the controlling questions are the same ones I laid out for the committee-stage bill, now with more weight behind them:

Provider or buyer? The tax falls on the provider of digital advertising services with in-state receipts. An agent platform that sells programmatic placement to Pennsylvania businesses looks like a provider. An agent that merely buys media for its own account looks like a buyer and is likely out of scope for the provider-level tax — though it still bears the underlying media's treatment. As more of the media-buying stack moves inside autonomous agents, that line is exactly what an auditor will test.

How are receipts sourced? Pennsylvania's gross-receipts approach inherits the same unresolved sourcing question every digital ad tax faces: is a programmatic receipt sourced to the advertiser, the viewer, or the server? Maryland has litigated this for years. Pennsylvania will have to answer it, and the answer determines how much of an agent platform's revenue lands in the Pennsylvania base.

Which template wins? This is now closer to settled than it was. The category-based gross-receipts model has the floor-vote durability the automation model lacks. If you are building tax logic for an agent that touches advertising, build it against the gross-receipts structure first — it is the one most likely to spread.

The Internet Tax Freedom Act Angle

One reason the gross-receipts model travels better is constitutional. The Internet Tax Freedom Act bars discriminatory taxes on electronic commerce. A base defined by "machine learning algorithms" or automated delivery invites a clean discrimination argument: it taxes the electronic method and arguably nothing else. A category-based gross-receipts tax on advertising providers can more plausibly sidestep that challenge by taxing the receipts, not the technology. Maryland's tax has been in litigation since 2021 and is still collecting. That track record — survivable, if contested — is precisely why a second state just adopted the same structure and pushed it through a floor vote.

What AgentTax Users Should Do Now

Separate provider receipts from buyer receipts. This is the substance an auditor tests under both the Pennsylvania and Maryland models. Tag receipts where your agent provides advertising services to third parties distinctly from receipts where it buys media.

Build a state-sourced digital-advertising receipts figure. Maryland, Illinois, and now Pennsylvania all want a clean, state-sourced number. Build it once; reuse it as the next state moves.

Track the Senate, not the headline. HB 1678 passed the House, not the Commonwealth. Model the exposure; do not remit speculatively on a bill that still has to clear the Senate and the Governor's desk by the end of the month.

Default to the gross-receipts template. When you design agent tax logic for advertising activity, assume the category-based structure is the one that proliferates. The June 10 vote is your evidence.

What to Watch

Three signals from here: whether the Pennsylvania Senate takes up HB 1678 before the June 30 budget deadline or lets it lapse; whether the Maryland appellate litigation validates or undermines the gross-receipts model both states now rely on; and whether any third state this cycle reaches for the Pennsylvania gross-receipts structure over the Illinois automation language. After June 10, the smart money is on the structure that just won 139-63.

If you operate agents that buy, place, or serve advertising and want to understand your multi-state exposure before these bills become law, that is exactly what we built AgentTax for — start with our AI agent tax compliance guide.

This analysis is for informational purposes only and does not constitute legal or tax advice. Consult a licensed tax professional for compliance decisions.