Claude AI users must now verify identity with government ID as Anthropic introduces new checks

Anthropic Introduces Identity Verification for Claude AI Users

Anthropic has started rolling out identity verification requirements for its Claude AI platform, asking users to provide a government-issued photo ID and live selfie when accessing certain capabilities 1

1

. The company announced the change this week, stating that verification prompts may appear during routine platform integrity checks or as part of safety and compliance measures 2

2

. While Anthropic hasn't publicly specified which features trigger verification, the company emphasizes that the process helps prevent abuse, enforce usage policies, and meet legal obligations 2

2

.

The verification process, which typically takes under five minutes, accepts physical documents from most countries including passports, driver's licenses, state or provincial ID cards, and national identity cards 2

2

. Photocopies, screenshots, digital IDs, student credentials, and temporary paper IDs are not accepted 2

2

. Users must also take a live selfie using their phone or webcam, which the system compares against the submitted ID 1

1

.

User Privacy Concerns Mount Over Data Handling

The announcement has generated significant backlash from users questioning why an AI chatbot requires identity verification, particularly for paying subscribers whose credit cards are already on file 1

1

. The timing proves especially awkward given that millions of users migrated to Claude AI from OpenAI in February after OpenAI signed a Pentagon deal to deploy AI on classified networks—a contract Anthropic declined over surveillance fears and concerns about autonomous weapons 3

3

. Daily signups broke records at the time, with free users up 60% since January as privacy-conscious individuals sought an alternative 3

3

.

Persona Identities Partnership Raises Questions

Anthropic selected Persona Identities as its verification partner, the same KYC infrastructure used across financial services and by companies like OpenAI and Roblox 1

1

3

3

. Criticism intensified when users discovered that Persona counts venture firm Founders Fund as a major investor—co-founded by Peter Thiel, who also co-founded and chairs Palantir, a surveillance company serving federal agencies including the FBI, CIA, and US Immigration and Customs Enforcement 1

1

. Palantir's services primarily expand government surveillance using facial recognition and AI technologies 1

1

.

Under the arrangement, Persona handles ID images and biometric data on its servers, not Anthropic's systems 3

3

. Anthropic functions as the data controller, setting rules for how information is used and stored, while Persona processes data on Anthropic's behalf 2

2

. Anthropic can access verification records through Persona's platform when needed, such as reviewing appeals, but does not copy or store those images 2

2

.

Data Security Promises and Historical Precedent

Anthropic has issued explicit assurances about data handling. The company states that verification data is used solely to confirm identity and meet legal and safety obligations, not for model training 2

2

. All data passing through the verification process is encrypted in transit and at rest 1

1

. Anthropic emphasizes it will not share identity data with third parties for marketing, advertising, or any purpose unrelated to verification and compliance, except where legally required to respond to valid legal processes 2

2

. Persona is contractually limited in how it can use the data, restricted to fraud prevention and verification improvement 3

3

.

However, third-party custody of government documents carries inherent risks. An October 2025 breach at Discord exposed roughly 70,000 government IDs that users had submitted for age verification 3

3

. While Persona maintains serious infrastructure, the incident demonstrates that no third party proves immune to security failures 3

3

.

Broader Implications for AI Accountability

The move fits a pattern Anthropic has been building toward greater accountability. In December, the company deployed classifiers to detect users who self-identify as minors, though multiple adult users reported account suspensions and lost project histories while appealing incorrect flags 3

3

. The verification system also enforces age restrictions, with under-18 access listed as grounds for account suspension 4

4

. Accounts registered from regions Anthropic doesn't formally serve face bans—a detail affecting Chinese users accessing the AI chatbot through intermediaries, since matching a live selfie against a physical government document proves difficult to circumvent 3

3

.

The shift signals that frictionless, consequence-free AI access is ending 4

4

. AI tools have become deeply embedded in daily work and life, and anonymity at scale creates potential for abuse and policy violations that simple email sign-ups cannot prevent 4

4

. Whether users find this development reassuring or intrusive likely depends on their usage patterns and trust in corporate data promises 4

4

. What remains clear is that platforms are introducing accountability measures for both users and providers, though reactions suggest many question whether "we promise not to misuse it" constitutes sufficient guarantee when handing government credentials to an organization built on consuming and processing data 4

4

.