Phoenix 24
The next technology race may be decided by practical fluency.
San Francisco, June 2026.
Anthropic has launched Claude Corps, a paid fellowship designed to train early-career professionals in the practical use of artificial intelligence while helping nonprofit organizations adopt the technology. Participants will receive an annual salary of $85,000 plus benefits during a year-long placement, even if they have no previous professional experience in AI. The initiative seeks to develop workers capable of integrating Claude into real organizational processes rather than limiting advanced tools to engineers and specialized laboratories. Anthropic has committed an initial $150 million to the program, which aims to train 1,000 fellows across several cohorts.
The first group will consist of 100 participants and is scheduled to begin in October 2026. Applications are expected to remain open until July 17, with candidates required to be at least 18 years old and possess fewer than two years of full-time professional experience. The program is therefore directed primarily toward people at the beginning of their careers rather than established technology executives. Its structure suggests that Anthropic views AI literacy as a foundational professional capability that can be developed through guided practice.
Fellows will not work directly as Anthropic employees. CodePath, a nonprofit organization focused on expanding access to technology careers, will serve as the official employer and manage salaries and benefits. Social Finance will evaluate the program’s results, including whether participants develop transferable skills and whether the host organizations achieve measurable improvements. This arrangement gives the initiative the characteristics of both a workforce-development program and a large-scale experiment in institutional AI adoption.
Participants will be assigned to nonprofit organizations across the United States, where they will identify operational problems that could be addressed with Claude. Their responsibilities may include automating repetitive administrative tasks, improving research processes, organizing internal knowledge, preparing documents or helping teams analyze information more efficiently. The fellowship is not designed to turn every participant into a machine-learning scientist. Instead, it focuses on creating specialists who understand how to apply existing AI systems responsibly within a specific professional environment.
At least 400 nonprofit organizations are expected to host fellows during the program. This focus addresses a persistent technological imbalance because charities, community institutions and public-interest organizations often lack the budgets necessary to recruit experienced AI engineers. Many also struggle to evaluate which tools are useful, safe or appropriate for sensitive information. By placing trained personnel inside these organizations, Anthropic can demonstrate practical applications of Claude while expanding its presence beyond corporate technology departments.
The $85,000 salary is one of the program’s most striking elements. Training initiatives frequently require participants to pay tuition, accept unpaid internships or assume personal financial risk while learning new skills. Claude Corps reverses that model by treating the acquisition and application of AI expertise as compensated professional work. The approach could attract candidates from economic backgrounds that traditionally make lengthy technical retraining difficult.
The program also reflects a broader change in the labor market. Companies increasingly need employees who can connect artificial intelligence with existing professional knowledge rather than specialists who only understand the underlying mathematics of large language models. A nonprofit worker familiar with education, healthcare, housing or community services may create more value by learning to use AI effectively than an engineer with little understanding of those sectors. This combination of domain expertise and technological fluency could become one of the most valuable employment profiles of the coming decade.
Anthropic also stands to benefit strategically. The fellows will introduce Claude into hundreds of organizations, generating examples of how the system performs in varied working conditions. Those experiences could reveal weaknesses, safety concerns and opportunities for product development while creating a network of professionals already familiar with the company’s technology. Although the initiative has a social mission, it also functions as a long-term investment in adoption and market influence.
Important questions remain about the nature of the expertise participants will acquire. Learning to prompt an AI system is not equivalent to understanding model architecture, data governance, cybersecurity or algorithmic bias. Effective training will therefore need to include verification practices, privacy protection and clear limits on when automated recommendations should not replace human judgment. Without those safeguards, organizations could become dependent on outputs they are not prepared to evaluate critically.
The use of AI within nonprofit institutions carries particular risks because these organizations may handle health records, financial information, immigration cases or data involving vulnerable communities. Fellows will need to distinguish between low-risk productivity tasks and decisions that require professional accountability. They must also understand that fluent responses can contain errors, fabricated references or misleading conclusions. The program’s credibility will depend on whether it teaches disciplined oversight rather than presenting AI as an automatic solution.
Claude Corps may become a model for other technology companies seeking to close the gap between rapid AI development and the slower transformation of the workforce. Paying people to learn while solving real organizational problems could prove more effective than relying exclusively on short online courses or theoretical certifications. It allows participants to build portfolios, develop professional networks and demonstrate measurable results. Host organizations, meanwhile, gain access to skills they might otherwise be unable to afford.
The initiative signals that the competition in artificial intelligence is expanding beyond more powerful models. The next stage will depend on who can train large numbers of people to use those systems effectively, safely and within the realities of everyday institutions. Anthropic is investing not only in Claude but also in the human infrastructure required to make the technology useful. In that emerging economy, expertise may belong less to those who merely understand AI and more to those who can translate it into responsible action.
Phoenix24: inteligencia para audiencias libres. / Phoenix24: intelligence for free audiences.