Full-Time
Develops reliable, interpretable AI systems
$240k - $315k/yr
H1B Sponsorship Available
San Francisco, CA, USA
Hybrid
In-office at least 25% of the time per policy; hybrid role in San Francisco office.
Bachelor's
See people who can refer or advise you
Anthropic focuses on AI research to build reliable, interpretable, and steerable AI systems. Its main product, Claude, is an AI assistant designed to handle tasks at any scale for clients across industries, delivered through deployment and licensing along with specialized AI R&D services. Claude works by combining natural language processing, human feedback, reinforcement learning, and interpretability techniques to produce a capable, controllable AI assistant that can assist with a wide range of tasks. The company differentiates itself from competitors by prioritizing safety, transparency, and controllability—emphasizing reliability, interpretability of model behavior, and user-controlled steerability in its AI systems. Anthropic’s goal is to make AI systems that people can trust and efficiently use to improve operations and decision-making across sectors.
Company Size
5,001-10,000
Company Stage
Series H
Total Funding
$162.8B
Headquarters
San Francisco, California
Founded
2021
See people who can refer or advise you
Help us improve and share your feedback! Did you find this helpful?
Flexible Work Hours
Paid Vacation
Parental Leave
Hybrid Work Options
Company Equity
Anthropic has hired Amir Salek, founder of Google's Tensor Processing Unit programme, for its compute team. Salek spent four years away from chip design after leaving Google in 2022 to invest in deep-tech startups at Cerberus Capital Management. This follows Anthropic's June hiring of Clive Chan, formerly the second engineer on OpenAI's custom chip programme. The AI company confirmed in August it is building an in-house silicon team to co-design chips with Claude models, aiming to halve inference costs. TechCrunch reported Anthropic has held early manufacturing discussions with Samsung. The moves come whilst Anthropic relies heavily on Google's TPUs, having agreed to purchase up to a million units in October.
Anthropic has introduced watermarks for text generated by its AI model Claude, following requirements under the European Union's Artificial Intelligence Act. The watermarks, which apply globally, use a method developed by Google that involves selecting specific words and fragments to create a statistically detectable pattern. The EU's AI Act, whose high-risk AI rules took effect this month, mandates transparency in AI-generated content. Companies violating the law face fines of up to 7% of global revenue. The watermarks are invisible to readers and don't affect text quality. They remain embedded when text is copied or edited, allowing potential identification of AI-generated content anywhere. The move is expected to increase tensions between the Trump administration and the EU over digital regulation. President Trump recently announced plans to review and potentially retaliate against what he called the EU's "discriminatory" digital practices.
Anthropic PBC has hired Amir Salek, a founder of Google's custom chip programme, as the AI lab prepares to develop its own semiconductors. Salek brings extensive experience from Alphabet's hardware efforts, where he played a key role in establishing the company's custom chip initiative. His appointment signals Anthropic's strategic shift towards hardware development. The move follows a broader industry trend of AI companies designing proprietary chips to optimise performance and reduce reliance on third-party suppliers. Custom semiconductors can offer significant advantages in power efficiency and computational capabilities for AI workloads. Anthropic joins other major AI labs in pursuing vertical integration through hardware development, a strategy that could provide greater control over its technology stack and long-term competitive positioning.
Anthropic will allow enterprise customers to store required data on their own cloud infrastructure, Reuters reported, citing an unnamed source. The AI company maintains its 30-day data retention requirement but now permits businesses to keep that data on their systems rather than Anthropic's. The policy applies to Anthropic's Fable and Mythos models and future frontier models. Anthropic introduced the 30-day retention requirement in June to protect against cyberattacks using its technology. The move follows mounting pressure on enterprise AI spending, with some companies switching to cheaper alternatives like DeepSeek. AI startup Lindy recently migrated from Anthropic's Claude models to DeepSeek, expecting to save millions of dollars within months. Anthropic filed a confidential IPO prospectus in June. Its annualised revenue run rate reached $47 billion in May, according to CNBC.
Medialister announced its MCP connector is now available in Anthropic's Claude Connectors Directory. The integration allows Claude users to research publications and identify editorial advertising opportunities through natural-language conversations within their existing AI workflow. Communications, SEO, content, and growth teams can use the connector to build media placement shortlists, compare outlets against campaign requirements, and turn briefs into research workflows inside Claude. Use cases include finding publications relevant to specific audiences, geographies, or industries, and planning product launches or PR campaigns. The connector is accessible through Claude's in-product browsing and search experience. Users can add it via Settings → Connectors → Browse Connectors. Whilst the connector handles discovery and planning, campaign booking remains within Medialister's platform. Medialister provides a platform for editorial advertising and media placements, helping brands and agencies discover publication opportunities and plan campaigns at scale.