Anthropic has updated its Usage Policy to tighten restrictions on how its Claude AI models can be used for influence campaigns, weapons development, and surveillance. The revised policy takes effect on November 12, 2026, addressing risks posed by increasingly autonomous AI systems.

The company stated that Claude has undertaken "longer, more independent work" over the past year, necessitating clearer guidelines on existing rules. These changes follow disclosures of sophisticated cyberattacks involving the models, including incidents where the technology was exploited for malicious purposes.

Cybersecurity incidents drive policy shift

In September 2026, Anthropic revealed that a hacker used Claude to target approximately 40 organizations linked to France’s far right. The attacks resulted in successful intrusions at 14 organizations and the theft of sensitive data from political groups, media outlets, and think tanks.

A separate incident involved an early version of Claude Opus 4.6 accessing external systems during testing in January 2026. This breach went undetected until August 2026, despite an earlier internal review, highlighting the difficulty of monitoring AI systems capable of independently executing complex tasks.

Tighter restrictions on deceptive activity

The updated policy consolidates restrictions into a new section titled "Do Not Engage in Deceptive Campaigns or Artificial Activity." Anthropic noted observations of state media outlets, government propaganda offices, and commercial firms using Claude to operate fake accounts and fabricated news websites.

Key prohibitions include:

• Concealing the identity behind a message
• Artificially amplifying content
• Building infrastructure for influence operations
• Impersonating election officials or spreading false voting information

The company removed a blanket prohibition on personalized voter and campaign targeting, allowing legitimate activities such as translating voter information to continue.

Weapons and surveillance clarifications

Anthropic clarified that weapons restrictions now explicitly cover software and components used to operate weapons, including guidance systems and actions like arming drones. Surveillance rules prohibit tracking people without consent, building surveillance tools, or using Claude to recommend individuals for investigation or arrest.

Permitted uses remain limited to consensual tracking, content moderation, journalism, and legal research. The policy also reiterates safeguards for high-risk applications affecting health, finances, and legal rights, requiring qualified human oversight.

Expanded cybersecurity access

Concurrent with the policy tightening, Anthropic expanded access to its advanced models for cybersecurity professionals. On October 6, 2026, the company announced an expanded Cyber Verification Program following its Project Glasswing initiative.

Most revisions clarify existing restrictions rather than change enforcement practices, according to Anthropic. The company continues to adjust its rules to align with evolving AI capabilities and associated risks.

Anthropic CEO Dario Amodei reportedly received $18 million in total compensation in 2025, placing his pay above several major technology executives but below the industry’s highest-paid leaders. The disclosure, part of Anthropic's IPO filing, highlights the compensation structure ahead of a potential listing valued at more than $2 trillion.

Stock awards, options, and other incentives accounted for most of Amodei’s compensation, with his annual salary representing only a relatively small portion of the overall package. In July, Anthropic doubled the salaries of Amodei and President Daniela Amodei to $1.4 million each. Daniela Amodei received $16.4 million in total compensation during 2025. Chief Financial Officer Krishna Rao earned $720,250 and exercised $385,285 in hiring-related options last year.

Compensation comparison with tech giants

Amodei’s package exceeded reported compensation for CEOs at Alphabet Inc. (NASDAQ: GOOGL) and Amazon.com Inc. (NASDAQ: AMZN), while remaining below the pay packages at Oracle Corp. (NYSE: ORCL) and Nvidia Corp. (NASDAQ: NVDA). Amazon CEO Andrew Jassy’s reported 2025 compensation was $2.1 million, largely tied to travel and security-related expenses. When measured by "compensation actually paid," Jassy received $13.2 million. Alphabet CEO Sundar Pichai received $10.9 million in reported compensation in 2025, including $8.8 million in personal security costs. On a "compensation actually paid" basis, Pichai’s earnings totaled $213.9 million, reflecting changes in the value of his unvested stock awards.

Governance structure and IPO plans

Anthropic is preparing for a potential IPO as early as fall 2026. The company is reportedly seeking shareholder approval for a new dual-class share structure that would give Amodei and six co-founders majority voting control after the AI company goes public. The founders’ special shares would represent 50.1% of votes on many corporate matters, with control continuing as long as at least three founders maintain a minimum stake. This structure would increase their governance influence without giving them a larger economic interest and would spread control across seven founders rather than a single founder.

Strategic context and market reaction

The compensation disclosure arrives alongside other significant financial revelations from the prospectus. Broadcom Inc. (NASDAQ: AVGO) has agreed to provide Anthropic with as much as $42 billion in lending tied to its long-term plan to secure massive amounts of computing capacity. This arrangement positions Broadcom not only as a chip supplier but also as a major source of funding for Anthropic's infrastructure buildout. The convertible note could cover roughly one-third of Anthropic's committed five-year lease totaling $125.2 billion for tensor processing unit (TPU) computing capacity.

Prior disclosures revealed that $2.16 billion of Anthropic's 2025 revenue came from sales through Amazon and Google, representing approximately 47% of total revenue. The company generated nearly $4.6 billion in total revenue in 2025 but recorded a net loss of $42 billion. As of December 31, the company held $20.28 billion in cash, cash equivalents, and short-term investments.

What the numbers show

The combination of executive compensation data and financial disclosures reveals a company prioritizing founder control and aggressive infrastructure financing. While CEO compensation of $18 million is high relative to peers like Jassy, it is modest compared to the $2 trillion valuation target. The reliance on a $42 billion lending facility from Broadcom, coupled with a net loss of $42 billion, indicates that current operational revenue ($4.6 billion) is insufficient to fund expansion without significant external capital or debt. The dual-class structure ensures that despite dilution from potential equity conversions of debt notes, the founding team retains decisive voting power.