> ## Content Index
> Fetch the complete content index at: https://readsludge.com/llms.txt
> Use this file to discover other available public pages before exploring further.

# AI Companies Are Funding the AGs Investigating Them
- URL: https://readsludge.com/2026/07/17/ai-companies-are-funding-the-ags-investigating-them/
- Published: 2026-07-17T18:42:45.000Z
- Updated: 2026-10-01T17:50:58.000Z
- Description: OpenAI, Google, Meta, and others sent a wave of donations to both parties’ attorney general groups as AGs move to regulate the AI industry. Elon Musk's xAI contributed as well.
- Author: David Moore
- Tags: Tech, DAGA, RAGA, attorneys general, state attorneys general, OpenAI, Google, Meta, xAI, AI

💸

**Share this on* [**X*](https://twitter.com/intent/tweet?url=https%3A%2F%2Freadsludge.com%2F2026%2F07%2F17%2Fai-companies-are-funding-the-ags-investigating-them%2F&text=AI%20Companies%20Are%20Funding%20the%20AGs%20Investigating%20Them) *and* [**Bluesky*](https://bsky.app/intent/compose?text=AI%20Companies%20Are%20Funding%20the%20AGs%20Investigating%20Them https%3A%2F%2Freadsludge.com%2F2026%2F07%2F17%2Fai-companies-are-funding-the-ags-investigating-them%2F)

Tech giants developing AI models—including OpenAI, Google, Meta, and Amazon—have donated more than $2 million combined to the national Democratic and Republican attorneys general campaign committees since the start of 2025, as AGs in both parties are suing AI companies over harms to children and deceptive safety claims.

More than half of that arrived in the past few months. Google, Meta, and Amazon each donated matching amounts in the second quarter of 2026—$250,000, $200,000, and $125,000, respectively—to the Democratic Attorneys General Association (DAGA) and the Republican Attorneys General Association (RAGA), according to the groups' IRS disclosures. 

The Democratic AG group received $25,000 in May from Elon Musk’s xAI, as the company is both suing California Attorney General Rob Bonta to block enforcement of a new AI transparency law—and is being investigated by Bonta this year over its Grok model's creation and distribution of sexually explicit deepfakes. Anthropic gave $250,000 to RAGA on May 6, its first donation to either group. 

In June, a coalition of state AGs led by New York Democrat Letitia James [subpoenaed](https://www.cnbc.com/2026/06/12/openai-says-its-engaging-constructively-with-state-ags-.html) OpenAI for documents on practices including treatment of minors and advertising, after user safety [incidents](https://www.pbs.org/newshour/nation/openai-hit-with-multistate-probe-into-possible-user-harm-as-its-ipo-looms). In August 2025, a bipartisan group of 44 state AGs—led by Republicans in South Carolina and Tennessee, alongside Democrats in Illinois and North Carolina—sent a [letter](https://www.naag.org/press-releases/bipartisan-coalition-of-state-attorneys-general-issues-letter-to-ai-industry-leaders-on-child-safety/) to OpenAI, Meta, Google, xAI, and other AI firms warning of "grave concerns" over children's interactions with AI chatbots. A follow-up [letter](https://www.attorneygeneral.gov/taking-action/ag-sunday-leads-coalition-of-42-attorneys-general-in-letter-to-a-i-software-companies-demanding-safeguards-to-protect-vulnerable-residents-from-harmful-interactions-with-bots/) in December, led by Pennsylvania Republican Dave Sunday and New Jersey Democrat Matthew Platkin, demanded stronger safeguards after harmful interactions with children and vulnerable users.

The DAGA donation from xAI was unexpected, given the company is in active litigation with one of the group’s prominent members. In December 2025, xAI [sued](https://news.bloomberglaw.com/ip-law/xai-sues-california-attorney-general-over-ai-training-data-law) Bonta in federal court, seeking to block AB 2013, a new California law requiring AI developers to disclose their training data. The company argued that the disclosure requirements amounted to an unconstitutional taking of trade secrets and compelled speech. Weeks later, on Jan. 16, Bonta [sent](https://www.reuters.com/legal/litigation/california-ag-sends-cease-desist-letter-xai-deepfake-images-2026-01-16/) xAI a separate cease-and-desist letter demanding that the company stop distributing sexually explicit deepfakes generated using Grok. Bonta said his office had opened an investigation after a wave of reports that Grok was being used to produce nonconsensual, sexualized images of women and girls—including content depicting minors—which he said met the legal definition of child sexual abuse material. He demanded xAI immediately halt the practice.

In March, a federal judge [denied](https://www.yahoo.com/news/articles/xai-loses-bid-halt-california-214208893.html) xAI's request to block the training-data transparency law while the suit continued. xAI then donated $25,000 on May 4—its first donation to either group.

### **States Move on AI Safety**

New AI safety laws are coming online in a few large Democratic-controlled states, and AGs are responsible for enforcing them. In California, [SB 53](https://calmatters.digitaldemocracy.org/bills/ca%5F202520260sb53), signed in September 2025, requires large developers of frontier AI models to document a safety framework and report any "critical safety incident" to the state within 15 days, among other things. The state AG can bring civil actions with penalties of up to [$1 million](https://leginfo.legislature.ca.gov/faces/billTextClient.xhtml?bill%5Fid=202520260SB53) per violation. In New York, the RAISE Act—signed in December 2025 and [finalized](https://www.wiley.law/alert-New-York-Finalizes-RAISE-Act-for-Frontier-AI-Models-Law-Takes-Effect-January-1-2027) in March 2026—similarly empowers the state AG to sue AI developers found to be noncompliant with safety-documentation duties, with penalties of up to [$1 million](https://www.governor.ny.gov/news/governor-hochul-signs-nation-leading-legislation-require-ai-frameworks-ai-frontier-models) for a first violation and up to $3 million for subsequent ones. New York’s law—which was severely [weakened](https://prospect.org/2025/12/11/hochul-caves-big-tech-ai-safety-bill-new-york/) by Gov. Kathy Hochul last year—will take effect on Jan. 1, 2027.

Republican AGs have moved to take on AI companies as well. In June, Florida’s James Uthmeier [sued](https://www.nbcnews.com/tech/tech-news/florida-sues-openai-sam-altman-saying-put-profit-safety-rcna347602) OpenAI and CEO Sam Altman, the first civil suit against the company by a state, alleging it concealed safety risks, especially to children and teenagers. In January, Kentucky's Russell Coleman [filed](https://www.kentucky.gov/Pages/Activity-stream.aspx?n=AttorneyGeneral&prId=1857) the first state lawsuit against an AI chatbot company, Character.AI, over harms to children. Last year, Texas AG Ken Paxton [opened an investigation](https://www.texasattorneygeneral.gov/news/releases/attorney-general-ken-paxton-investigates-meta-and-characterai-misleading-children-deceptive-ai) into Meta AI Studio and Character.AI, looking into marketing of chatbot products as mental health tools in ways that could be used by children and vulnerable people.

In the absence of a federal framework governing AI, more than half of U.S. states have enacted around 109 AI laws and 28 data center laws in 2026, according to a [tally](https://www.techpolicy.press/where-state-ai-legislation-stands-half-way-into-2026/) by the NYU Center on Technology Policy, on top of [around 100](https://www.ncsl.org/technology-and-communication/artificial-intelligence-2025-legislation) AI measures adopted or enacted in 2025\. Watchdog groups [say](https://readsludge.com/2025/12/12/dem-governors-gathered-with-ai-lobbyists-ahead-of-trump-preemption-eo/) states are stepping in to fill the vacuum with rules ranging from consumer protections to bans on algorithmic discrimination and more.

### **Dollars to DAGA and RAGA**

Google gave $250,000 to DAGA in June, according to the group’s second-quarter [filing](https://forms.irs.gov/app/pod/basicSearch/downloadFile?formId=167556&formType=e8872), after giving $50,000 in November, and its sibling company Waymo gave $25,000 in April. Amazon gave $125,000 in June, adding to $30,000 it gave in October and $125,000 the previous June. Meta donated $200,000 in April after a $5,000 contribution in November.

OpenAI has donated $200,000 to DAGA this cycle, giving $75,000 in April 2025—its first donation to the group—and $125,000 in March 2026\. Other AI interests have also given to DAGA since the start of 2025: Microsoft contributed $52,500 and Oracle $50,000, each in a pair of donations.

Google's June donation was a big step up from the two $25,000 donations it gave DAGA in 2020, according to IRS contribution data compiled by [ProPublica](https://projects.propublica.org/527-explorer/orgs/134220019/). Donations from Meta, Microsoft, and Amazon were comparable to past years.

On the Republican side, Google, Meta, and Amazon matched their DAGA gifts to RAGA in the [second quarter](https://forms.irs.gov/app/pod/basicSearch/downloadFile?formId=167408&formType=e8872) of this year, adding to [donations](https://forms.irs.gov/app/pod/basicSearch/downloadFile?formId=160302&formType=e8872) [made](https://forms.irs.gov/app/pod/basicSearch/downloadFile?formId=164790&formType=e8872) by Google and Amazon to RAGA in 2025\. Anthropic gave [$250,000](https://forms.irs.gov/app/pod/basicSearch/downloadFile?formId=167408&formType=e8872) on May 6, its first RAGA donation. Previously, Oracle chipped in [$25,000](https://forms.irs.gov/app/pod/basicSearch/downloadFile?formId=165899&formType=e8872) to RAGA in March, Flock gave [$25,000](https://forms.irs.gov/app/pod/basicSearch/downloadFile?formId=164790&formType=e8872) in July 2025, and OpenAI gave [$75,000](https://forms.irs.gov/app/pod/basicSearch/downloadFile?formId=160302&formType=e8872) in May 2025.

OpenAI’s dealings with Democratic AGs goes back to before this year’s AI safety laws. The company needed sign-off from two of them last year to complete a weighty restructuring: in October, Bonta [approved](https://calmatters.org/economy/technology/2025/10/openai-restructuring-deal-full-of-holes-critics-say/) OpenAI's conversion toward a for-profit company, along with Delaware Attorney General Kathy Jennings, after pressing the company for changes first. The restructuring drew protest: OpenAI's move away from nonprofit sparked an [open letter](https://notforprivategain.org/) of opposition from former employees and legal experts, who argued for the importance of nonprofit control over the company's goal of building artificial general intelligence (AGI). After the agreement with Bonta was reached, former OpenAI safety lead Steven Adler [told](https://calmatters.org/economy/technology/2025/10/openai-restructuring-deal-full-of-holes-critics-say/) CalMatters that the company's safety committee should have greater independence to be effective.

Neither DAGA nor RAGA responded to a request for comment on their AI industry funding, in light of their members pursuing cases with donors like xAI, OpenAI, and Meta.

### **Data Center Battles**

DAGA's website says its members are focused on affordability, even as its AI industry donors are building, or buying power from, massive data centers [contributing](https://thehill.com/policy/technology/5970522-data-center-power-costs-pjm/) mightily to rising energy prices. 

In December, DAGA [announced](https://dems.ag/icymi-new-effort-announced-from-state-ags-to-protect-consumers/) a Consumer Protection and Affordability Working Group under its policy affiliate, aimed at strengthening consumer protections and countering the Trump administration's rollback of industry rules. The group was tasked with recommending strategies in areas including technology, health care, and financial services. It was led by Rohit Chopra, former director of the Consumer Financial Protection Bureau (CFPB), who in May took on a consumer protection role in California’s government.

Data centers’ immense energy consumption has become a flashpoint from local boards to statehouses to Congress; Gallup recently [found](https://news.gallup.com/poll/709772/americans-oppose-data-centers-area.aspx) that seven in ten Americans oppose data center construction locally. In New York, Hochul just [signed](https://gothamist.com/news/hochul-puts-hold-on-large-data-centers-in-new-york) an executive order establishing a one-year moratorium on construction of large data centers using more than 50 megawatts of electricity while the state writes new rules, after the legislature passed a bill setting a one-year pause on facilities using 20 megawatts.

With data centers' appetites [driving up](https://www.pbs.org/newshour/economy/massive-data-center-buildout-poses-latest-inflation-threat-for-consumers) energy prices, at least [six states](https://finance.yahoo.com/sectors/energy/articles/electric-bills-rise-states-focusing-195811598.html) have recently challenged utility rate increases hitting residents. Arizona's Kris Mayes, a Democrat up for re-election, is trying to block rate hikes before the state's utility board, saying, "I felt like it's never been more important to stand up against the blatant corporate greed of our monopoly utilities in Arizona." North Carolina's Jeff Jackson, also a Democrat, [urged](https://www.wunc.org/2026-06-08/consumer-advocates-push-nc-state-regulators-to-create-separate-rates-for-data-centers) the state utility commission to create separate rates for large consumers like data centers to shield ratepayers. Earlier this year, Michigan Democrat Dana Nessel similarly [called for](https://www.bloomberg.com/news/articles/2026-02-05/michigan-ag-urges-revisit-on-oracle-data-center-power-plan) reconsideration of approval for utility DTE Energy to feed power to a data center campus planned by Oracle and OpenAI, both DAGA donors.

OpenAI is also linked with spending campaigns at the federal level seeking to knock out candidates who support AI safety rules, like New York Assemblymember Alex Bores, co-author of the RAISE Act. The company’s president Greg Brockman and his wife Anna have [bankrolled](https://readsludge.com/2026/04/09/ai-super-pacs-are-unleashing-millions-to-tilt-primaries-in-their-favor/) the Leading the Future super PAC network with $25 million as it [pushes](https://www.cnbc.com/2026/07/09/ai-companies-election-spending.html) for [federal preemption](https://readsludge.com/2026/02/24/ai-boom-on-k-street-one-in-four-lobbyists-now-work-on-ai/) of state AI laws through a national framework, one aligned with the Trump administration’s [executive order](https://www.whitehouse.gov/presidential-actions/2025/12/eliminating-state-law-obstruction-of-national-artificial-intelligence-policy/) on AI last year. (A few OpenAI employees [donated $245,000](https://thehill.com/policy/technology/5971905-openai-eployees-245k-ai-safety-pac/) to a safety-focused super PAC, Guardrails Alliance, that seeks to counter Leading the Future.) Meta committed [$65 million](https://www.politico.com/news/2026/02/02/meta-drops-65-million-into-super-pacs-to-boost-tech-friendly-state-candidates-00759567) to two new spending groups in state elections backing candidates it deems as supportive, one focused on California races.

🎁

**Help keep our two-person nonprofit newsroom on this Big Money beat.* [**Give a gift subscription*](https://readsludge.com/gift-subscription/) *to Sludge.*