Anthropic (Claude)

Quick Take

Anthropic has made government rules for AI the center of its politics. It endorsed California's SB 53 AI safety bill in 2025, opposed a proposed 10-year federal ban on state AI laws, and has given $40 million to Public First Action, a group that pushes for AI safeguards.[8][9][10] That stance drew fire from the Trump White House, whose AI adviser David Sacks accused the company in October 2025 of "a sophisticated regulatory capture strategy based on fear-mongering."[11] Anthropic's employee-funded PAC, launched in April 2026, has split its federal giving almost evenly between Republicans and Democrats.[12][13] In California, where companies may give directly, state records show Anthropic gave $850,000 to the California Democratic Party in August 2026.[14] Federal lobbying rose from $720,000 in 2024 to $3.13 million in 2025 and $5.64 million in the first nine months of 2026.[15]

The Fight With the Pentagon

Anthropic's sharpest clash with Washington is over how the military may use Claude. Anthropic says it was the first frontier AI company to put its models on the government's classified networks and has supported the military since June 2024.[3] In July 2025 the Defense Department's AI office gave it a two-year agreement worth up to $200 million to build AI tools for national security work.[16]

The terms Anthropic set for that work became the problem. The company said it would support all lawful national security uses except two: mass domestic surveillance of Americans and fully autonomous weapons, meaning weapons that pick and strike targets with no human decision.[3] It said today's AI is not reliable enough to run such weapons and that mass surveillance of Americans violates fundamental rights.[3] The Pentagon's position, as NPR reported it, was that a contractor should not tell the military how it can use a product.[17]

When talks broke down, the government acted on two fronts on the same day. President Trump posted that he was directing "EVERY Federal Agency" to stop using Anthropic's technology, with a six-month phase-out for the Pentagon and other agencies that already relied on it.[2] Hegseth announced on X that he was directing his department to designate Anthropic a supply chain risk, a label that can bar a company from defense work.[3] Anthropic said the label had been aimed at U.S. adversaries in the past and had never been publicly used against an American company. It said, "No amount of intimidation or punishment from the Department of War will change our position."[3]

The department confirmed the designation in a letter on March 4, 2026. Anthropic said it would fight it in court and offered to keep supplying its models to the military at nominal cost during any transition.[18] In the same statement, Amodei apologized for the tone of an internal message to employees that had leaked to the press.[18] According to The Information, as quoted by Forbes, that message said the administration did not like Anthropic in part because "we haven't donated to Trump (while OpenAI/Greg have donated a lot)," a reference to OpenAI and its president, Greg Brockman.[5]

The court fight has produced split results. Anthropic sued in both California and Washington, D.C.[19] On August 27, 2026, U.S. District Judge Rita Lin in California ruled the designation was unlawful retaliation that violated the First Amendment, writing that "the empty invocation of national security is not a blank check."[19] On September 25, 2026, a federal appeals court in Washington ruled 2 to 1 the other way in the second case. Judges Gregory Katsas and Neomi Rao held that a 2018 federal supply chain law let the Pentagon exclude Claude from military systems; Judge Karen LeCraft Henderson dissented.[20] The Congressional Research Service, which advises Congress, says the Pentagon may keep removing Anthropic products from its systems while review continues.[1] In October, Anthropic asked the full appeals court to rehear the case.[21]

The Lobbying Footprint

Anthropic's federal lobbying has grown every year, and it now spends more than $2 million a quarter. In 2023 it lobbied only through an outside firm, Aquia Group, and OpenSecrets, a nonpartisan group that tracks money in politics, puts that year at $280,000.[22][15] Anthropic began filing its own lobbying reports in 2024 and reported $720,000 that year.[15][22]

Spending then jumped. Anthropic's own reports to Congress total $3.13 million for calendar year 2025. They show $1.56 million in the first quarter of 2026, $1.97 million in the second, and $2.11 million in the third, for $5.64 million through September 30.[15] The third-quarter figure was a record for the company and about double its biggest quarter of 2025.[23] In 2026 it added Ballard Partners, Navigators Global and Avenue Solutions to the outside lobbying firms it already used.[15] Axios reported that the company's second-quarter 2026 filings covered export controls, cybersecurity, AI safety standards, Defense Department purchasing, funding for the Energy Department's Genesis Mission, and AI in health care. Its lobbyists met with lawmakers in both chambers and with officials at the White House, the Commerce Department and the Treasury Department.[24]

Its third-quarter 2026 report lists work on AI bills, export controls and "distillation" (copying a model by training a rival on its answers), AI provisions in the defense budget bill, AI provisions in the Kids Online Safety Act, and energy rules for data centers in the Ratepayer Protection Act.[6][23] Lobbying reports list the topics a company worked on, not whether it supported or opposed each one.[25]

Pushing for AI Rules, Including in the States

Anthropic has backed binding safety rules for the most powerful AI systems, including state laws, while saying it would prefer one national standard. In August 2024, Amodei told California Governor Gavin Newsom that an amended version of SB 1047, a state AI safety bill, was improved to the point that "its benefits likely outweigh its costs," though he said Anthropic was "not certain of this."[26]

In September 2025 Anthropic endorsed SB 53, a narrower California bill. It requires the largest AI developers to publish how they manage catastrophic risks, report serious safety incidents to the state within 15 days, and protect whistleblowers, with fines for breaking their own published commitments.[8] Companies with less than $500 million in yearly revenue are exempt, a carve-out Anthropic backed to protect startups.[9] Anthropic said frontier AI safety is best handled by the federal government but that advances in AI will not wait for Washington to agree.[8] In December 2025 it published its plan for complying with the law, which it said would take effect January 1. Anthropic said the law balanced its safeguards with flexibility in how companies carry them out.[27]

At the federal level, Anthropic opposed a proposed 10-year moratorium that would have blocked state AI laws; the Senate voted 99 to 1 against the moratorium.[9] Amodei wrote that Anthropic favors a single federal standard and wants to work with both parties to write one.[9]

The White House pushed back. In October 2025, Sacks wrote that Anthropic was "principally responsible for the state regulatory frenzy that is damaging the startup ecosystem."[11] A week later Amodei answered in a public statement. He pointed to Anthropic's federal contracts, said the company had publicly praised President Trump's AI Action Plan and supported the administration's push to expand energy for AI, and wrote that managing AI's effects "should be a matter of policy over politics."[9]

The Company, Its Leaders, and Their Money

Anthropic's federal PAC splits its giving almost evenly between the parties. The company's largest political gift in California went to the California Democratic Party.[14]

The company PAC. Anthropic registered the Anthropic PBC Political Action Committee, known as AnthroPAC, with the Federal Election Commission on April 3, 2026.[12] A PAC (political action committee) is a fund a company sets up to give to candidates; this one is paid for by voluntary employee donations, not company money.[28][29] Through October 14, 2026, it raised $403,937 and gave $320,750 to candidates, party committees and other PACs.[12]

RedRightBuyer counted every one of those gifts in the PAC's federal reports. $160,000 went to Republicans and $152,750 to Democrats, with $8,000 to three leadership PACs whose party we did not confirm.[13] Its largest gifts were $15,000 each to the four national House and Senate campaign committees of both parties.[13] A leadership PAC is a fund run by a sitting politician; four of the ones AnthroPAC gave to were assigned to a party here based on the candidates they themselves fund. The Hill, using spending records shared with it ahead of the federal filing deadline, reported a similar split for July through September: 50.6 percent to Republicans and 49.4 percent to Democrats.[29]

Outside spending. Anthropic has also put $40 million, in two $20 million gifts in February and July 2026, into Public First Action.[30][10] Anthropic says the money goes only to the group's public education and policy work and "cannot be used to influence the election of any candidate."[10] Public First was created by two former congressmen, Republican Chris Stewart of Utah and Democrat Brad Carson of Oklahoma.[31] Super PACs tied to Public First, Jobs and Democracy and Defending Our Values, had spent $16.6 million on congressional races by June 2026, according to OpenSecrets data reported by NPR.[32] One of them, Defending Our Values, spent nearly $880,000 backing Republican Rep. Celeste Maloy in a Utah primary.[31]

These groups sit across from Leading the Future, a super PAC network funded mainly by the venture firm Andreessen Horowitz and by Brockman, which opposes stricter AI rules.[32] The two sides collided in a New York City Democratic primary, where groups linked to OpenAI and Anthropic spent more than $23 million for and against Alex Bores, a state lawmaker who co-sponsored New York's AI safety law. Bores finished second.[32]

California. Federal databases do not cover state politics, and in California Anthropic gives directly as a company. State records show Anthropic itself gave $850,000 to the California Democratic Party in two gifts on August 5, 2026, of $801,000 and $49,000. It also gave $500,000 on September 30, 2026, to the campaign for Proposition 1, a statewide veterans and affordable housing bond; $12,500 to the Bay Area Legislative Leaders PAC; and $207,400 to state candidates.[14] The San Francisco Chronicle reported that more than $100,000 of the candidate money went to Democrats running for the state Senate and Assembly.[33]

The founders. Amodei personally gave $1 million on May 4, 2026, to Public First, a super PAC (a committee that can spend unlimited sums on elections but cannot coordinate with candidates), according to federal records.[34] The Campaign Legal Center, a nonpartisan watchdog, reported that the super PAC had raised about $4.9 million, with $1 million from Public First Action, $1 million from Amodei, and most of the rest from other Anthropic employees.[35] Forbes, citing OpenSecrets, reported that Dario and Daniela Amodei appear to be registered Democrats who have given thousands of dollars to Democratic and independent candidates and none to Republicans.[5] These are the founders' own donations, not company money.

Investors. Amazon has invested $8 billion in Anthropic.[36] Anthropic's September 2025 funding round was led by the investment firm ICONIQ, with Fidelity and Lightspeed Venture Partners. It also included two government-owned investment funds, Singapore's GIC and the Qatar Investment Authority.[37] These investors' own political records are not treated here as Anthropic's.

Employees. Before AnthroPAC existed, federal giving tied to Anthropic came only from individuals. OpenSecrets counts $9,200 in the 2024 election cycle from people tied to the company, $8,900 of it to Democrats and the rest to a liberal political group, with Sen. Jacky Rosen and Kamala Harris the top recipients.[22] OpenSecrets reports no outside spending by Anthropic in that cycle.[22]

The Policy Agenda in Its Own Words

Beyond safety rules, Anthropic's stated agenda puts heavy weight on national security and competition with China. Anthropic says it is the only frontier AI company that refuses to sell its services to companies controlled by the People's Republic of China, at a cost of significant revenue. It has backed limits on selling advanced U.S. chips to China.[9] It also calls for tighter export controls on chips and chipmaking equipment and for measures against distillation attacks.[10] It has also said there should be channels for reporting model distillation attacks to a designated government agency.[38]

On safety, Anthropic says publishing safety plans is not enough. It wants the government able to check safety claims, impose civil penalties, and slow or block release of models that pose serious catastrophic risk.[10] Its policy proposals say frontier AI developers should test their models, be open about the results and submit them to independent evaluation.[38] Its policy page says a technology moving this fast "should not be governed by the industry alone."[39]

On energy, Anthropic's July 2025 energy report said the U.S. AI industry needs at least 50 gigawatts of new electric capacity by 2028. It called for faster permitting for geothermal, natural gas and nuclear power and for new power line buildouts.[40] In February 2026 the company said it would pay 100 percent of the grid upgrades needed to connect its own data centers and would work to cover consumer electricity price increases tied to them. It also said it supports federal permitting reform.[41]

On jobs, it has published what it calls an Economic Policy Framework, an early plan for responding to AI-driven disruption of the labor market.[39] In June 2025 it launched an Economic Futures Program offering research grants of up to $50,000 to study AI's effects on work, along with policy events in Washington and London.[42]

Anthropic also says it trains Claude to be politically even-handed, giving opposing views "equal depth, engagement, and quality of analysis." In November 2025 it published a test of that claim in which its own models scored 94 to 95 percent on its even-handedness measure.[43] That is the company's description of its product and its own test of it.

Other Dealings With the Government

Anthropic has worked with both recent administrations on AI safety. In July 2023 it was one of seven companies, with Amazon, Google, Inflection, Meta, Microsoft and OpenAI, that made voluntary safety commitments to the Biden White House, including outside security testing of models before release.[44] In August 2024 it signed an agreement giving the government's U.S. AI Safety Institute access to its major new models before and after public release; the Trump administration re-established that office in June 2025 as the Center for AI Standards and Innovation.[45] In March 2025, weeks after the change in administration, Anthropic removed the 2023 White House commitments from its online Transparency Hub, including a pledge to research AI bias and discrimination, according to the watchdog group The Midas Project. Co-founder Jack Clark said the removal was not a change in substance and that Anthropic continues to follow the commitments.[46]

Before the Pentagon dispute, Anthropic was also expanding its government business, including Claude Gov, a version built for government and national security customers.[16] In August 2025 it agreed, through the General Services Administration's OneGov program, to offer Claude to all three branches of the federal government for $1 for one year.[47]

In June 2026 the government restricted one of Anthropic's own products. On June 12, citing national security, it issued an export control directive barring foreign nationals from using two new models, Claude Fable 5 and Claude Mythos 5. Anthropic could not check users' nationality in real time, so it shut both models off for everyone and called the order a misunderstanding.[48][49] Anthropic said the controls were lifted on June 30, and Fable 5 returned on July 1. It said the government's Center for AI Standards and Innovation had tested its safeguards.[49]

Outside the United States

Anthropic's political engagement abroad has centered on AI rules in Europe and Britain. In July 2025 it said it would sign the European Union's General-Purpose AI Code of Practice, a voluntary guide to complying with the EU's AI Act, saying the code "advances the principles of transparency, safety and accountability."[50] In February 2025 it signed a memorandum of understanding with the British government to explore using Claude in public services and to supply testing tools to the UK's AI Security Institute.[51] RedRightBuyer did not retrieve any EU lobbying register figures for Anthropic for this profile.

How Anthropic Is Owned and Run

Anthropic is a public benefit corporation, a type of for-profit company that must also pursue a stated public purpose. Its stated purpose is "the responsible development and maintenance of advanced AI for the long-term benefit of humanity."[52] Its board is elected by stockholders and by the Long-Term Benefit Trust, an independent body whose members hold no financial stake in the company.[7][52] When it announced the trust in 2023, Anthropic said the trust would elect a majority of the board within four years.[52] The trustees listed on its website include former Federal Reserve chairman Ben Bernanke.[7]

Anthropic is privately held, so it files no annual report with a workforce or diversity section. Its main careers page says its hiring process looks for "thoughtful people with diverse expertise" and does not use the words equity or inclusion.[53] A posting for its Fellows Program says people from underrepresented groups are more prone to imposter syndrome and that Anthropic strives "to include a range of diverse perspectives on our team."[54]

Where to Find Them

Claude is available at claude.ai on the web, desktop and phone, and businesses use it through Anthropic's developer platform. Company information is at anthropic.com.

Footnotes

[1] Congressional Research Service, CRS Insight IN12669, "Pentagon-Anthropic Dispute over Autonomous Weapon Systems: Potential Issues for Congress," March 13, 2026, updated October 5, 2026. https://www.everycrsreport.com/reports/IN12669.html [2] AP and Reuters, Al Jazeera, "Trump orders federal agencies to stop using Anthropic as dispute escalates," February 27, 2026. https://www.aljazeera.com/news/2026/2/27/trump-orders-federal-agencies-to-stop-using-anthropic-as-dispute-escalates [3] Anthropic, Anthropic News, "Statement on the comments from Secretary of War Pete Hegseth," February 27, 2026. https://www.anthropic.com/news/statement-comments-secretary-war [4] Anthropic, Anthropic News, "Anthropic Raises Series B to build steerable, interpretable, robust AI systems," April 29, 2022. https://www.anthropic.com/news/anthropic-raises-series-b-to-build-safe-reliable-ai [5] Richard Nieva, Forbes, "The Anthropic Cofounders Standing Up To Trump," March 4, 2026. https://www.forbes.com/sites/richardnieva/2026/03/04/the-anthropic-cofounders-standing-up-to-trump/ [6] Anthropic, U.S. Senate Lobbying Disclosure, "Third Quarter 2026 Lobbying Report," filed October 20, 2026. https://lda.gov/filings/public/filing/d0b9f104-2212-4fcc-8304-5c978c366768/print/ [7] Anthropic, Anthropic, "Company," accessed October 24, 2026. https://www.anthropic.com/company [8] Anthropic, Anthropic News, "Anthropic is endorsing SB 53," September 8, 2025. https://www.anthropic.com/news/anthropic-is-endorsing-sb-53 [9] Dario Amodei, Anthropic News, "A statement from Dario Amodei on Anthropic's commitment to American AI leadership," October 21, 2025. https://www.anthropic.com/news/statement-dario-amodei-american-ai-leadership [10] Anthropic, Anthropic News, "Donating another $20 million to Public First Action," July 21, 2026. https://www.anthropic.com/news/donation-public-first-action [11] Bryan Pietsch, Bloomberg Law, "Anthropic Is Running a Strategy Based on Fear-Mongering: Sacks," October 14, 2025. https://news.bloomberglaw.com/artificial-intelligence/anthropic-is-running-a-strategy-based-on-fear-mongering-sacks [12] Federal Election Commission, FEC.gov, "Anthropic PBC Political Action Committee (C00946111): committee overview," accessed October 24, 2026. https://www.fec.gov/data/committee/C00946111/?cycle=2026 [13] Federal Election Commission, FEC.gov, "Anthropic PBC Political Action Committee: regularly filed reports, May through October 2026" (Schedule B itemized disbursements, tallied by RedRightBuyer), accessed October 24, 2026. https://www.fec.gov/data/committee/C00946111/?tab=filings [14] California Secretary of State, Power Search, "Contributions by Anthropic" (data current as of October 23, 2026), accessed October 24, 2026. To reproduce: open Advanced Search, enter contributor "Anthropic," all dates, and download the CSV (335 records). The California Democratic Party rows (committee ID 1018392) show Anthropic PBC contributions of $801,000 and $49,000 dated August 5, 2026. https://powersearch.sos.ca.gov/advanced.php [15] U.S. Senate Office of Public Records, Lobbying Disclosure Act database, "Filings for client Anthropic, 2023 to 2026" (all years, paginated; 2026 filings on later pages), accessed October 24, 2026. https://lda.gov/api/v1/filings/?client_name=Anthropic [16] Anthropic, Anthropic News, "Anthropic and the Department of Defense to advance responsible AI in defense operations," July 14, 2025. https://www.anthropic.com/news/anthropic-and-the-department-of-defense-to-advance-responsible-ai-in-defense-operations [17] Geoff Brumfiel, NPR, "Judge says the Pentagon can't designate AI company Anthropic a 'supply chain risk'," August 28, 2026. https://www.npr.org/2026/08/28/nx-s1-5947951/judge-says-the-pentagon-cant-designate-ai-company-anthropic-a-supply-chain-risk [18] Dario Amodei, Anthropic News, "Where things stand with the Department of War," March 5, 2026. https://www.anthropic.com/news/where-stand-department-war [19] Rebecca Bellan, TechCrunch, "Anthropic gets its first court win over the Pentagon's supply-chain risk label," August 28, 2026. https://techcrunch.com/2026/08/28/anthropic-gets-its-first-court-win-over-the-pentagons-supply-chain-risk-label/ [20] Alexandra Agraz, Law Commentary, "Anthropic Loses Appeals Court Bid to Overturn Pentagon 'Supply Chain Risk' Label," September 25, 2026. https://www.lawcommentary.com/articles/anthropic-loses-pentagon-supply-chain-risk-appeal [21] Emilia David, BankInfoSecurity, "Anthropic Seeks Full Appeals Court Review of Pentagon Ban," October 14, 2026. https://www.bankinfosecurity.com/anthropic-seeks-full-appeals-court-review-pentagon-ban-a-33105 [22] OpenSecrets, OpenSecrets.org, "Anthropic PBC Profile: Summary," accessed October 24, 2026. https://www.opensecrets.org/orgs/anthropic-pbc/summary?id=D000106114 [23] Annemarije de Boer, Yahoo Finance, "OpenAI, Anthropic, and Nvidia Spent Combined $7.5 Million On Lobbying in Q3," October 22, 2026. https://finance.yahoo.com/technology/ai/articles/openai-anthropic-nvidia-spent-combined-191134517.html [24] Maria Curi, Axios, "Anthropic ramps up lobbying spending amid AI policy fights," July 21, 2026. https://www.axios.com/2026/07/21/anthropic-ramps-up-lobbying-spending-ai-policy-fights [25] Ashley Gold, Semafor, "Meta tops Big Tech lobbying as AI labs talk antitrust," October 21, 2026. https://www.semafor.com/article/10/21/2026/meta-tops-big-tech-lobbying-as-ai-labs-talk-antitrust [26] John K. Waters, Campus Technology, "Anthropic Announces Cautious Support for New California AI Regulation Legislation," August 26, 2024. https://campustechnology.com/articles/2024/08/26/anthropic-announces-cautious-support-for-new-california-ai-regulation-legislation.aspx [27] Anthropic, Anthropic News, "Sharing our compliance framework for California's Transparency in Frontier AI Act," December 19, 2025. https://www.anthropic.com/news/compliance-framework-SB53 [28] Jason Nelson, Decrypt, "AI Giant Anthropic Files to Launch 'AnthroPAC' Amid Clash With Trump Administration," April 4, 2026. https://decrypt.co/363355/ai-giant-anthropic-anthropac-clash-trump-administration [29] Miranda Nazzaro and Julia Shapero, The Hill, "Anthropic PAC splits Q3 donations down the middle," October 15, 2026. https://thehill.com/homenews/6152693-anthropic-pac-splits-q3-donations-down-the-middle/amp/ [30] Maria Curi, Axios, "Anthropic pours $20 million into AI policy fight," February 12, 2026. https://www.axios.com/2026/02/12/anthropic-millions-ai-policy-fight [31] Brigham Tomco, Deseret News (via KSL.com), "The national AI policy showdown is being decided in Utah's backyard," June 13, 2026. https://www.ksl.com/article/news/politics/the-national-ai-policy-showdown-is-being-decided-in-utahs-backyard/51509173 [32] Shannon Bond and Eric McDaniel, NPR, "AI and tech are trying to influence the midterm elections," June 22, 2026, updated June 24, 2026. https://www.npr.org/2026/06/22/nx-s1-5856359/ai-anthropic-congress-spending-openai-midterms-election [33] Hanna Zakharenko and Roland Li, San Francisco Chronicle, "OpenAI and Anthropic execs donate millions to political causes. One billionaire stands apart," October 5, 2026. https://www.sfchronicle.com/california/article/openai-anthropic-political-donation-22456454.php [34] Federal Election Commission, FEC.gov, "Individual contributions: Dario Amodei to Public First (C00930503)," receipt dated May 4, 2026, accessed October 24, 2026. https://www.fec.gov/data/receipts/individual-contributions/?contributor_name=dario+amodei&two_year_transaction_period=2026&min_amount=100000 [35] Campaign Legal Center, Campaign Legal Center, "The AI Industry's Influence Operation: How Big Money Undermines Democracy, and What Can Be Done About It," September 29, 2026. https://campaignlegal.org/document/ai-industrys-influence-operation-how-big-money-undermines-democracy-and-what-can-be-done [36] Anthropic, Anthropic News, "Powering the next generation of AI development with AWS," November 22, 2024. https://www.anthropic.com/news/anthropic-amazon-trainium [37] Anthropic, Anthropic News, "Anthropic raises $13B Series F at $183B valuation," September 2, 2025. https://www.anthropic.com/news/anthropic-raises-series-f-at-usd183b-post-money-valuation [38] Anthropic, Anthropic, "Policy on the AI Exponential," accessed October 24, 2026. https://www.anthropic.com/policy-on-the-ai-exponential [39] Anthropic, Anthropic, "AI policy," accessed October 24, 2026. https://www.anthropic.com/policy [40] Anthropic, Anthropic News, "Build AI in America: Anthropic Energy Report," July 21, 2025. https://www.anthropic.com/news/build-ai-in-america [41] Zachary Skidmore, Data Center Dynamics, "Anthropic to cover consumer electricity price increases tied to its data center energy consumption," February 13, 2026. https://www.datacenterdynamics.com/en/news/anthropic-to-cover-consumer-electricty-price-increases-tied-to-its-data-center-energy-consumption/ [42] Anthropic, Anthropic News, "Introducing the Anthropic Economic Futures Program," June 27, 2025. https://www.anthropic.com/news/introducing-the-anthropic-economic-futures-program [43] Anthropic, Anthropic News, "Measuring political bias in Claude," November 13, 2025. https://www.anthropic.com/news/political-even-handedness [44] Joseph R. Biden, Jr., The American Presidency Project, "FACT SHEET: Biden-Harris Administration Secures Voluntary Commitments from Leading Artificial Intelligence Companies to Manage the Risks Posed by AI," July 21, 2023. https://www.presidency.ucsb.edu/node/363710 [45] National Institute of Standards and Technology, NIST News, "U.S. AI Safety Institute Signs Agreements Regarding AI Safety Research, Testing and Evaluation With Anthropic and OpenAI," August 29, 2024. https://www.nist.gov/news-events/news/2024/08/us-ai-safety-institute-signs-agreements-regarding-ai-safety-research [46] Habiba Katsha, POCIT, "Anthropic Quietly Removes Commitments To Tackle AI Bias And Discrimination From Website," March 6, 2025. https://peopleofcolorintech.com/articles/anthropic-quietly-removes-commitments-to-tackle-ai-bias-and-discrimination-from-website/ [47] Alexandra Kelley, Nextgov/FCW, "GSA and Anthropic ink deal for Claude AI across all government branches," August 12, 2025. https://nextgov.com/artificial-intelligence/2025/08/gsa-and-anthropic-ink-deal-claude-ai-across-all-government-branches/407377/ [48] Anthropic, Anthropic News, "Statement on the US government directive to suspend access to Fable 5 and Mythos 5," June 12, 2026. https://www.anthropic.com/news/fable-mythos-access [49] Anthropic, Anthropic News, "Redeploying Fable 5," June 30, 2026, updated July 1, 2026. https://www.anthropic.com/news/redeploying-fable-5 [50] Anthropic, Anthropic News, "Anthropic to sign the EU Code of Practice," July 21, 2025. https://www.anthropic.com/news/eu-code-practice [51] Ingrid Lunden, TechCrunch, "UK drops 'safety' from its AI body, now called AI Security Institute, inks MOU with Anthropic," February 13, 2025. https://techcrunch.com/2025/02/13/uk-drops-safety-from-its-ai-body-now-called-ai-security-institute-inks-mou-with-anthropic/ [52] Anthropic, Anthropic News, "The Long-Term Benefit Trust," September 19, 2023. https://www.anthropic.com/news/the-long-term-benefit-trust [53] Anthropic, Anthropic, "Careers," accessed October 24, 2026. https://www.anthropic.com/careers [54] Anthropic, Anthropic job board (Greenhouse), "Anthropic Fellows Program, AI Safety & Security," accessed October 24, 2026. https://job-boards.greenhouse.io/anthropic/jobs/5183044008