Anthropic Public Record — key findings
Nationally representative survey of 51,993 Americans, fielded Nov–Dec 2025 (YouGov)
TCS deal, $15M cyber program, and a trust deficit: Anthropic brief, June 12
Three material events from Friday, June 12: Anthropic signed a global enterprise deal with TCS (50,000 employees, 56 countries), launched a $15 million cyber defense program for state and local governments, and published the first results of its 52,000-person public survey — which found that only 15% of Americans trust AI companies.
Three material events landed on Anthropic's newsroom on Friday, June 12: a large enterprise services deal with TCS, a new government cyber defense program, and the first results from a 52,000-person public opinion survey on AI.
TCS partnership: 50,000 employees, 56 countries
Category: Customer
Anthropic announced a partnership with Tata Consultancy Services (TCS), one of the world's largest IT services firms. 1
Under the agreement, TCS will deploy Claude to 50,000 of its own employees across 56 countries, and build industry-specific Claude-powered products for clients in financial services, healthcare, life sciences, aviation, telecom, and the public sector. TCS joins the Claude Partner Network.
The deal includes four active use cases already underway:
- Diligenta (TCS's UK life and pensions subsidiary) will use Claude for customer experience improvements serving 22 million policyholders
- TCS's banking and financial services product teams will use Claude Code for software engineering and IT operations
- Engineering teams will build reusable Claude Code plugins for claims adjudication and lending advisory
- TCS iON, which conducts more than 75 million assessments per year across 1,500 cities in India, will deliver Claude training and certification
Dario Amodei said the deal deepens Anthropic's commitment to India, describing it as the company's "second-largest market."
$15M cyber defense program for state and local governments
Category: Customer / Product
Anthropic launched a $15 million cyber defense program targeted at state, local, tribal, and territorial (SLTT) governments — a population left with funding gaps after federal programs including the State and Local Cybersecurity Grant Program were cancelled under the Trump administration. 2
The program provides up to $100,000 in Claude credits per eligible smaller government. Larger entities — special districts managing power, water, transit, or port infrastructure — may qualify for additional slots. Credits can be applied toward Claude Opus 4.8, Claude Security, and Claude Code, along with cyber defense runbooks and workflow templates.
California and Texas joined the first cohort, which kicked off Thursday. Cohorts will continue on a rolling basis through summer 2026. California's chief information security officer, Vitaliy Panych, and Texas Cyber Command Chief TJ White (VADM, ret.) both issued statements endorsing the initiative.
Anthropic distinguished this program from Project Glasswing, its restricted-access defensive cybersecurity initiative that gives select researchers access to Claude Mythos Preview. The SLTT program uses publicly available models only.
Anthropic Public Record: first results from a 52,000-person survey
Category: Research / Policy signal
Anthropic published the first results from its Anthropic Public Record, a nationally representative YouGov survey of 51,993 Americans fielded November–December 2025. 3
統計カードを読み込んでいます…
The trust gap is stark. Among institutions tested, AI companies ranked last at 15% — below the federal government (20%), state and local government (19%), international bodies (20%), and well below independent experts (43%).
チャートを読み込んでいます…
Bipartisan support for AI regulation stands at 71% overall — 79% among Democrats, 68% among Republicans, 69% among Independents — with a majority in every state. The top desired areas: privacy (56%), child safety (52%), and liability for harm (49%).
The gap between AI users and non-users is real but not in the direction one might expect: Americans who use AI at work every day are 16 percentage points less worried about cognitive dependency, and 16 points less worried about job displacement, compared to those who never use it. Anthropic notes that hands-on experience may reveal AI's limitations as well as its augmentation potential.
The survey is methodologically notable for its geographic granularity: state samples range from n=232 (Alaska) to n=1,902 (New York), weighted to Census benchmarks. Anthropic says the Public Record will be repeated regularly and expanded outside the US.
The results were announced alongside Anthropic's Economic Policy Framework and Advanced AI Framework — the company's paired policy proposals on AI governance and labor market preparation published June 10.
このコンテンツについて、さらに観点や背景を補足しましょう。