Section 230 turned 30 and Congress is debating its sunset, AI data centers are straining the power grid faster than it can be upgraded, and there's still no federal data privacy law after three failed Congresses — how much should Washington regulate the digital infrastructure everything else now runs on?
Each issue breaks into the specific questions Congress actually fights over. Read each position, then head to the interactive version of this issue to mark which reflects your view and build a message to your representatives.
Congress is under real pressure to act now that juries, not just legislators, are weighing in on algorithmic harm. Meta and other platforms faced significant jury verdicts in 2026 tied to algorithmic content decisions, intensifying pressure on Congress to narrow Section 230's liability shield. Bipartisan sunset legislation — including a bill that would end Section 230 protection entirely after December 31, 2026 absent replacement legislation — reflects genuine, cross-party frustration that the law now shields algorithmic curation decisions its authors never anticipated.
Everyone agrees the law needs reform; almost no one agrees on what it should actually be replaced with. The bipartisan breadth of support for reform — co-sponsors spanning from Josh Hawley to Amy Klobuchar to Lindsey Graham — masks real disagreement about what should replace Section 230's current protections. Conservatives largely want platforms held accountable for biased content moderation and viewpoint suppression, while many Democrats want accountability for algorithmic amplification of harmful content — meaning "reform Section 230" means substantively different things depending on who's saying it.
Removing the shield would mainly help the biggest platforms survive compliance costs that would crush smaller competitors. Big technology companies spent tens of millions of dollars lobbying against Section 230 changes in early 2026 alone. Critics of reform — including smaller platform coalitions like Internet Works — warn that removing the liability shield would primarily benefit the largest, best-resourced platforms, while crushing smaller competitors and startups the original 1996 law was designed to protect.
Chasing the lowest sticker price eliminated the training programs that determine whether new connections actually get used. NTIA's 2025 restructuring of the $42.45 billion BEAD program toward a technology-neutral "Benefit of the Bargain" bidding process eliminated dedicated funding for the digital-literacy and workforce-training programs that determine whether newly connected rural households actually adopt and benefit from broadband service. That prioritizes the lowest sticker price per connection over the full picture of what makes a broadband rollout succeed long-term.
The program is moving fast now; whether that speed cost too much in technology quality is the live question. The restructuring genuinely accelerated a program that had been stalled for years — more than 50 of 56 eligible entities had Final Proposal approval by mid-2026. Whether that speed came at an acceptable cost in technology quality and program scope remains contested, since "technology-neutral" in practice means substituting fixed-wireless or satellite for fiber in some areas originally slated for fiber-optic connections.
Taxpayers shouldn't overpay for one specific technology when a cheaper one delivers comparable service faster. Requiring costly fiber buildouts in every location, as the original BEAD rules effectively did, was driving per-location costs so high that many rural communities faced years of additional delay. Organizations representing fixed-wireless providers argue the technology-neutral approach lets rural areas get connected faster and at lower public cost.
A $720 billion transmission bill risks landing on ordinary ratepayers instead of the companies building the AI infrastructure driving it. U.S. data centers are projected to consume 9-17% of national electricity generation by 2030 — more than double current levels — and FERC's June 2026 order directing regional grid operators to speed data-center interconnection prioritizes tech-industry demand. Consumer advocates warn the resulting transmission upgrades, estimated at $720 billion nationally, risk being passed through to ordinary ratepayers' electric bills rather than absorbed by the companies building the AI infrastructure.
The federal order only handles interconnection speed — whether costs land on ratepayers is really a state-by-state question now. FERC's order explicitly states it acts to prevent cost-shifting onto other transmission customers while leaving retail rate protection to the states. Whether AI's power costs land on ordinary electricity customers depends heavily on decisions state utility regulators haven't all made yet, not on the federal order itself, which only addresses how quickly data centers can physically connect to the grid.
The grid's supply problem predates AI and needs faster investment regardless — slowing data centers down doesn't fix that. PJM's December 2025 capacity auction failed for the first time in its history to procure enough power to meet reliability targets, a genuine warning sign that the grid needs faster generation and transmission investment regardless of AI's specific role. Streamlining interconnection — rather than blocking or slow-walking data-center growth — is the more direct fix for a supply problem that predates the current AI buildout.
Even the newly restored authority comes with a congressional minimum and deadline attached, not full independent discretion. The FCC's general spectrum auction authority lapsed in 2023 and was only narrowly restored via the One Big Beautiful Bill Act, which now directs a much larger 160 MHz Upper C-Band auction the FCC approved on July 22, 2026. Congress attached that restoration to a specific minimum (100 MHz) and timeline (by July 2027) rather than handing the FCC full, permanent discretion, meaning the agency's spectrum pipeline still depends on Congress setting the terms.
The FCC moved unusually fast on paper, but real-world deployment still won't happen for years regardless. The FCC's July 22, 2026 order authorizing the Upper C-Band auction — 160 MHz, 60 MHz above the legal minimum — shows the restored authority working faster than many expected, moving from proposal to order in record time. The auction's winning bidders still won't be able to deploy new service until December 2030 after required aviation-safety retrofits, showing that spectrum policy remains a multi-year process regardless of how quickly the FCC itself acts.
Restoring authority with clear conditions let the FCC move faster than an open-ended grant might have, while still protecting aviation safety. The FCC's July 22, 2026 vote to auction 160 MHz of Upper C-Band spectrum — exceeding the 100 MHz minimum Congress required — demonstrates that restoring auction authority through targeted legislation, paired with close coordination with the FAA on aviation safety, can move fast. The auction is projected to raise billions of dollars for the Treasury while expanding the contiguous spectrum block available for next-generation wireless service.
Every federal preemption bill on the table would strip away the private-lawsuit option that gives the strongest state privacy laws their teeth. Nearly 20 states now have their own comprehensive data-privacy laws, several with private rights of action letting individuals sue companies directly. Every federal preemption bill introduced in 2026 — including the Consumer Data Privacy and Security Act — would override those state laws in favor of a single federal standard enforced only by the FTC and state attorneys general, eliminating the private lawsuit option.
The same two sticking points have sunk federal privacy legislation three Congresses running, despite real bipartisan interest in the goal. The core disagreement in every federal privacy bill introduced since 2024, including 2026's competing proposals, has consistently been the same two questions — should federal law preempt stronger state laws like California's, and should individuals be able to sue directly. No version has resolved both in a way that keeps both major parties' negotiators on board, which is why comprehensive federal privacy legislation has now failed in three consecutive Congresses despite genuine bipartisan interest in the goal.
A confusing 20-state patchwork is a real burden smaller businesses struggle with, and agency enforcement can have real teeth too. A single national privacy standard would replace a genuinely confusing 20-state patchwork of different consumer rights, opt-out mechanisms, and compliance deadlines that smaller businesses in particular struggle to navigate. FTC-and-state-AG enforcement — the model in 2026's Consumer Data Privacy and Security Act — provides real regulatory teeth without the litigation costs and inconsistent outcomes that come from allowing private lawsuits under 50 different potential legal standards.