# Pew Research Center: Political Ads on Facebook and Instagram Criticize Data Centers [Canonical reading](https://escapemeridian.com/readings/pew-research-center-political-ads-on-facebook-and-instagram-6a1e491b) Anna Lieb, a computational social science analyst at Pew Research Center, reports that U.S. political ads mentioning data centers on Facebook and Instagram have become more critical during 2026. In October, 57% of these ads opposed data centers, while 3% supported them. In January, 31% opposed them and 42% supported them. Pew analyzed U.S. ads about social issues, elections or politics that mentioned data centers in text or audio and ran from January 1 to October 24, 2026. Pew finds that ads opposing data centers more often discuss corporate interests, while supportive ads more often discuss jobs. Advertisers spent almost $12 million on these ads from July 27 to October 24, Pew estimates, using the midpoint of Meta’s reported spending ranges. These figures describe advertising, not public opinion or the effects of the ads. Pew used GPT-5.6 Luna to classify ads; on a validation dataset of 300 human-coded ads, the model had an average F1 score of 0.88 and 95% accuracy. How closely do these advertising trends reflect changes in voters’ views of data centers? Headline and summary by Escape Meridian, prepared by Luna from the attributed source. A source check is not independent confirmation of the source’s claims. ## Original source Pew Research Center · Article Original source date: Oct 29, 2026 (2026-10-29T12:53:50-04:00). Date basis: Date stated by the publisher. Added to Escape Meridian: 2026-11-07T00:46:32.466Z. First observed: 2026-10-29T16:54:46.674Z. Retrieved: 2026-10-29T17:02:22.310Z. - [On Facebook and Instagram, Political Ads Are Increasingly Critical of Data Centers](https://www.pewresearch.org/short-reads/2026/10/29/on-facebook-and-instagram-political-ads-are-increasingly-critical-of-data-centers/) ## Limits The analysis covers U.S. ads on Facebook and Instagram that mentioned data centers in text or audio and ran from January 1 to October 24, 2026; it does not establish how representative these ads are of all political messaging or public opinion. Meta provides ad spending data as a range, and Pew used the midpoint of each range to calculate total spending. The model’s reported F1 score and accuracy come from a validation dataset of 300 human-coded ads; these measures do not establish real-world reliability beyond that evaluation. ## Open question How closely do these advertising trends reflect changes in voters’ views of data centers? ## Source-stated Pew Research Center analysis found that, among political ads mentioning data centers, opposition was more common than support in October 2026. > So far in October, 57% of these ads have expressed opposition to data centers, while just 3% have been clearly in favor of them. - [Supporting source](https://www.pewresearch.org/short-reads/2026/10/29/on-facebook-and-instagram-political-ads-are-increasingly-critical-of-data-centers/) ## Source-stated In January 2026, support for data centers was more common than opposition in the ads Pew analyzed. > 42% were supportive of them while 31% were critical. - [Supporting source](https://www.pewresearch.org/short-reads/2026/10/29/on-facebook-and-instagram-political-ads-are-increasingly-critical-of-data-centers/) ## Source-stated Ads opposing data centers discussed corporate interests more often than supportive ads did; supportive ads discussed jobs more often. > nearly half of ads that oppose data centers discuss the corporate interests behind them - [Supporting source](https://www.pewresearch.org/short-reads/2026/10/29/on-facebook-and-instagram-political-ads-are-increasingly-critical-of-data-centers/) ## Source-stated Pew estimated almost $12 million in spending on political ads mentioning data centers from July 27 to October 24, using the midpoint of Meta’s ad spend ranges. > advertisers on Facebook and Instagram spent almost $12 million on ads that mention data centers - [Supporting source](https://www.pewresearch.org/short-reads/2026/10/29/on-facebook-and-instagram-political-ads-are-increasingly-critical-of-data-centers/) ## Source-stated Pew used GPT-5.6 Luna to classify ads, and reports its average F1 score and accuracy on a validation dataset coded by trained human coders. > the model achieved an average F1 score of 0.88 and 95% accuracy on a validation dataset of 300 ads coded by trained human coders. - [Supporting source](https://www.pewresearch.org/short-reads/2026/10/29/on-facebook-and-instagram-political-ads-are-increasingly-critical-of-data-centers/) ## Source and contributor context Anna Lieb · Article author These are credits stated by the source; they do not establish a profile match or authorship of every passage. ## Source media Media declared by the original publisher; availability checked separately from the editorial reading. - [Pew Research Center illustration · Pew Research Center](https://www.pewresearch.org/wp-content/uploads/sites/20/2026/10/SR_26.10.29_data-center-ads_featured.jpg?resize=1600,900) ## Terms and references Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement. ## Pew Research Center Pew Research Center studies public attitudes and social change through surveys and other empirical research. - [Explore this context](/entities/pew-research-center) - [Reference](https://www.pewresearch.org/) - [Reference](https://www.pewresearch.org/internet/2025/04/03/how-the-us-public-and-ai-experts-view-artificial-intelligence/) ## Editorial provenance Version 1. Publication ID: b7573517-22ba-40ed-82c9-2c18b4a0d1c2. Prepared by gpt-6-luna; grounding review by gpt-6-luna. Source check: passed by anthropic/claude-opus-5.5 on 2026-11-07T00:46:32.466Z. This checks fidelity to the source, not whether its claims are independently true. Editorial selection: accept; reasons: useful\_perspective. Observation: 28c65b8eb26404dcf69e1be0451f7122842d30ba42311c11592c729fc255567a. Source text hash: c88e3da17d5dc4ae17ae2630d884454a90c1479812e05640d05c2b49276b26c2. ## Revision and source-check history ### Version 1 — 2026-11-07T00:46:32.466Z Anna Lieb, a computational social science analyst at Pew Research Center, reports that U.S. political ads mentioning data centers on Facebook and Instagram have become more critical during 2026. In October, 57% of these ads opposed data centers, while 3% supported them. In January, 31% opposed them and 42% supported them. Pew analyzed U.S. ads about social issues, elections or politics that mentioned data centers in text or audio and ran from January 1 to October 24, 2026. Pew finds that ads opposing data centers more often discuss corporate interests, while supportive ads more often discuss jobs. Advertisers spent almost $12 million on these ads from July 27 to October 24, Pew estimates, using the midpoint of Meta’s reported spending ranges. These figures describe advertising, not public opinion or the effects of the ads. Pew used GPT-5.6 Luna to classify ads; on a validation dataset of 300 human-coded ads, the model had an average F1 score of 0.88 and 95% accuracy. How closely do these advertising trends reflect changes in voters’ views of data centers? ## Original source Pew Research Center · Article Original source date: Oct 29, 2026 (2026-10-29T12:53:50-04:00). Date basis: Date stated by the publisher. Added to Escape Meridian: 2026-11-07T00:46:32.466Z. First observed: 2026-10-29T16:54:46.674Z. Retrieved: 2026-10-29T17:02:22.310Z. - [On Facebook and Instagram, Political Ads Are Increasingly Critical of Data Centers](https://www.pewresearch.org/short-reads/2026/10/29/on-facebook-and-instagram-political-ads-are-increasingly-critical-of-data-centers/) ## Limits The analysis covers U.S. ads on Facebook and Instagram that mentioned data centers in text or audio and ran from January 1 to October 24, 2026; it does not establish how representative these ads are of all political messaging or public opinion. Meta provides ad spending data as a range, and Pew used the midpoint of each range to calculate total spending. The model’s reported F1 score and accuracy come from a validation dataset of 300 human-coded ads; these measures do not establish real-world reliability beyond that evaluation. ## Open question How closely do these advertising trends reflect changes in voters’ views of data centers? ## Source-stated Pew Research Center analysis found that, among political ads mentioning data centers, opposition was more common than support in October 2026. > So far in October, 57% of these ads have expressed opposition to data centers, while just 3% have been clearly in favor of them. - [Supporting source](https://www.pewresearch.org/short-reads/2026/10/29/on-facebook-and-instagram-political-ads-are-increasingly-critical-of-data-centers/) ## Source-stated In January 2026, support for data centers was more common than opposition in the ads Pew analyzed. > 42% were supportive of them while 31% were critical. - [Supporting source](https://www.pewresearch.org/short-reads/2026/10/29/on-facebook-and-instagram-political-ads-are-increasingly-critical-of-data-centers/) ## Source-stated Ads opposing data centers discussed corporate interests more often than supportive ads did; supportive ads discussed jobs more often. > nearly half of ads that oppose data centers discuss the corporate interests behind them - [Supporting source](https://www.pewresearch.org/short-reads/2026/10/29/on-facebook-and-instagram-political-ads-are-increasingly-critical-of-data-centers/) ## Source-stated Pew estimated almost $12 million in spending on political ads mentioning data centers from July 27 to October 24, using the midpoint of Meta’s ad spend ranges. > advertisers on Facebook and Instagram spent almost $12 million on ads that mention data centers - [Supporting source](https://www.pewresearch.org/short-reads/2026/10/29/on-facebook-and-instagram-political-ads-are-increasingly-critical-of-data-centers/) ## Source-stated Pew used GPT-5.6 Luna to classify ads, and reports its average F1 score and accuracy on a validation dataset coded by trained human coders. > the model achieved an average F1 score of 0.88 and 95% accuracy on a validation dataset of 300 ads coded by trained human coders. - [Supporting source](https://www.pewresearch.org/short-reads/2026/10/29/on-facebook-and-instagram-political-ads-are-increasingly-critical-of-data-centers/) ## Source and contributor context Anna Lieb · Article author These are credits stated by the source; they do not establish a profile match or authorship of every passage. ## Source media Media declared by the original publisher; availability checked separately from the editorial reading. - [Pew Research Center illustration · Pew Research Center](https://www.pewresearch.org/wp-content/uploads/sites/20/2026/10/SR_26.10.29_data-center-ads_featured.jpg?resize=1600,900) ## Terms and references Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement. ## Pew Research Center Pew Research Center studies public attitudes and social change through surveys and other empirical research. - [Explore this context](/entities/pew-research-center) - [Reference](https://www.pewresearch.org/) - [Reference](https://www.pewresearch.org/internet/2025/04/03/how-the-us-public-and-ai-experts-view-artificial-intelligence/) ## Editorial provenance Version 1. Publication ID: b7573517-22ba-40ed-82c9-2c18b4a0d1c2. Prepared by gpt-6-luna; grounding review by gpt-6-luna. Source check: passed by anthropic/claude-opus-5.5 on 2026-11-07T00:46:32.466Z. This checks fidelity to the source, not whether its claims are independently true. Editorial selection: accept; reasons: useful\_perspective. Observation: 28c65b8eb26404dcf69e1be0451f7122842d30ba42311c11592c729fc255567a. Source text hash: c88e3da17d5dc4ae17ae2630d884454a90c1479812e05640d05c2b49276b26c2. ### passed — 2026-11-07T00:46:32.466Z — anthropic/claude-opus-5.5 Publication: b7573517-22ba-40ed-82c9-2c18b4a0d1c2 attribution — pass: The reading correctly attributes the analysis to Anna Lieb of Pew Research Center. It presents the figures as Pew's findings and estimates, not as independent facts. > Anna Lieb is a computational social science analyst at Pew Research Center. summary — pass: The reading leads with the shift from January (42% supportive, 31% critical) to October (57% critical, 3% supportive). It states the scope and method, the corporate-interests vs. jobs themes and the spending estimate. The title fairly reflects that most recent ads are critical. The question asks an open issue rather than asserting an answer. A minor optional gap: it omits Pew's note that the work builds on its data center and AI research and its finding that candidates of both parties are similarly critical. Neither omission prevents understanding. F1 score is a technical term suited to a tooltip. > But since the beginning of this year, ads that mention data centers have grown increasingly critical. date — pass: The originalPublished value of 2026-10-29 matches the article's dateline. All dated figures (January, October, Jan. 1 to Oct. 24, July 27 to Oct. 24) match the source. > Short Reads | October 29, 2026 support — pass: All claims match the source: the percentages, the theme differences including jobs in supportive ads, the almost $12 million estimate based on range midpoints, and the model's validation metrics. The limitations correctly separate advertising data from public opinion and test metrics from real-world reliability. No benefits or costs are implied beyond the evidence. > Meanwhile, ads that support data centers disproportionately discuss their impact on jobs. media — pass: The selected media is the verified link to the original Pew article. It is the most direct way to reach the source. ## Recorded references Explicit citations in public readings and public file attachments to this reading. Shared topics are not citations. Other artifact types need an explicit reference adapter. No public references recorded here yet. [Reference data (JSON)](/readings/6a1e491b83eb894e83d90aeaeeae88d3f95b2dc11c68d11c4457b1d7b4298e64/references.json) ## Corrections and challenges [Send a correction or an opposing source](https://escapemeridian.com/corrections/) [JSON record](https://escapemeridian.com/readings/pew-research-center-political-ads-on-facebook-and-instagram-6a1e491b.json)