Full-Time
Associate General Counsel
AI Product & Privacy
Arena Intelligence
51-200 employees
Crowdsourced LLM evaluation and comparison platform
No salary listed
California, USA
Hybrid
Hybrid work is based in the Bay Area.
JD
 Legal (1) |
|---|
Get referred to Arena Intelligence
See people who can refer or advise you
- The candidate must have 8 or more years of legal experience in privacy and product counseling at a top-tier law firm and/or in-house legal department, ideally with experience counseling on consumer products.
- The candidate must have a Juris Doctor degree and active membership in at least one United States state bar, with California preferred, in good standing.
- The candidate must have independently built and run a global privacy program, with deep hands-on familiarity with the General Data Protection Regulation, California Consumer Privacy Act, California Privacy Rights Act, and cross-border data transfer mechanisms.
- The candidate must be able to advise product and engineering teams in real time and translate legal requirements into practical, build-ready guidance.
- The candidate must be able to think strategically and in detail to drive complex problem-solving and develop creative, business-forward solutions.
- The candidate must be able to calibrate risk, distinguish business risk from legal risk, and communicate it clearly to executive, product, and technical audiences.
- The candidate must be able to work collaboratively across the organization on projects of varying size.
- The candidate must be able to think quickly in a fast-paced environment and demonstrate strong organizational and communication skills, particularly under launch and regulatory deadlines.
- The candidate must have a deep interest in artificial intelligence, generative artificial intelligence, and machine learning technology, with the ability to engage effectively with technical teams on data architecture and evaluation issues.
- The candidate must be able to build and iterate on personal tools and create human-in-the-loop workflows that scale legal advice and work product, with experience using tools such as Cursor, Claude Code, Replit, Lovable, and/or Retool.
- Serve as the embedded legal partner to product and engineering teams, advising on privacy regulation, intellectual property, enterprise use cases, consumer protection, and related matters from concept through launch.
- Own Arena's global privacy program, including compliance with the General Data Protection Regulation, California Consumer Privacy Act, California Privacy Rights Act, and the broader international privacy landscape, with an emphasis on privacy by design and product architecture.
- Advise on the responsible development of artificial intelligence products and emerging regulations related to artificial intelligence model, agent, and tool development and deployment.
- Help shape how the company navigates novel questions at the intersection of privacy, artificial intelligence, and human evaluation.
- Design and maintain the contractual and operational frameworks governing the company's data flows.
- Negotiate privacy provisions and agreements with customers, partners, and vendors.
- Build and scale privacy and product legal processes, create playbooks, and implement systems to support a high-growth organization.
- Build agents and internal software to support legal operations.
- Procure legal technology tools and integrate internally developed and vendor-built software.
- Experience advising on consumer products or multi-sided marketplaces and building global privacy programs focused on controller responsibility.
- Experience advising on de-identification and anonymization data pipelines.
Arena Intelligence provides a crowdsourced, open platform for evaluating large language models (LLMs) by running head-to-head comparisons of two anonymous models on user prompts. Users vote on the better response and, after voting, model identities are revealed; the results generate large-scale preference data used to compute Elo ratings published on a public leaderboard. Originating from LMSYS and UC Berkeley’s Sky Computing Lab as Chatbot Arena, the project evolved into a company and rebranded to Arena, with $100 million seed funding in 2024 to scale the platform. The underlying framework, FastChat, remains open source, including the user interface and serving backend. Arena differentiates itself by providing a transparent, real-world benchmarking method that aggregates community-driven judgments and offers analytics to AI labs and enterprises, aiming to guide model development with observable performance data.
Company Size
51-200
Company Stage
Series A
Total Funding
$250M
Headquarters
San Francisco, California
Founded
2025
Get referred to Arena Intelligence
See people who can refer or advise you
Simplify's Take
What believers are saying
- Arena reached $100 million annualized revenue in June 2026, eight months after launch.
- A January 2026 Series A raised $150 million at a $1.7 billion valuation.
- The March 2026 leaderboard changelog added direct-battle votes, increasing volume and improving ranking stability.
What critics are saying
- 2025 research exposed provider-specific sampling, score retraction, and private-variant gaming on Arena.
- If enterprises trust distorted rankings, Arena’s neutrality brand breaks, crippling AI Evaluations sales in 2026.
- Scale AI, Mercor, and Surge compete directly on evaluation workflows and can undercut pricing.
What makes Arena Intelligence unique
- Arena’s leaderboard shapes model selection across OpenAI, Google, Anthropic, and DeepSeek.
- Its 10-million-plus human comparisons create a live preference dataset competitors cannot replicate.
- Arena’s AI Evaluations monetizes that trust with enterprise-grade analytics and audits.
Help us improve and share your feedback! Did you find this helpful?
Benefits
Health Insurance
Dental Insurance
Vision Insurance
Company Equity
Growth & Insights and Company News
6 month growth
↓ -8%1 year growth
↓ -3%2 year growth
↓ -9%Companies are struggling to decide which AI models to trust, according to Arena AI CEO Anastasios Angelopoulos. Speaking on the "20VC" podcast, he said enterprises face a difficult choice between frontier labs and Chinese open-source providers. Frontier companies like OpenAI and Anthropic offer advanced models but charge premium prices and can shut off access without warning or impose safeguards limiting capabilities. Chinese firms like DeepSeek offer cheaper open-weight models that run locally, but face potential security concerns and possible US restrictions. Arena operates a crowdsourced platform where users compare AI models. In June, the company announced it had reached $100 million in annualised run-rate revenue within eight months of launching its enterprise service. Angelopoulos said some model providers have tried using Arena's internal data to get employees to label data, though he declined to name specific companies.
Arena, the AI leaderboard provider that began as a UC Berkeley research project in 2023, has reached $100 million in annualised run-rate revenue just eight months after launching its commercial service. The company's post-money valuation stands at $1.7 billion following a $150 million Series A round in January. Arena is known for its crowdsourced AI model performance leaderboard, generated from over 10 million user evaluations. Whilst the public leaderboard remains free, the company monetises through AI Evaluations, a service providing model labs and enterprises with performance analytics. The startup competes with human labelling companies like Mercor, Surge and Scale AI for post-training refinement services. Arena has raised $250 million from investors including Felicis, Andreessen Horowitz and Kleiner Perkins. Co-founders include CEO Anastasios Angelopoulos, CTO Wei-Lin Chiang and Databricks co-founder Ion Stoica.
Arena, formerly LM Arena, has been valued at $1.7 billion just seven months after launching as a UC Berkeley PhD research project. The startup operates the leading public leaderboard for frontier large language models, influencing funding decisions and product launches across the AI industry. Co-founders Anastasios Angelopoulos and Wei-Lin Chiang claim their platform is harder to manipulate than static benchmarks, using what they call "structural neutrality" to evaluate AI models. The company is backed by the very companies it ranks, including OpenAI, Google and Anthropic. Arena currently shows Claude topping expert leaderboards in legal and medical use cases. The startup is expanding beyond chat to benchmark AI agents, coding and real-world tasks through a new enterprise product.
LMArena achieves $1.7B valuation four months after launching its product. LMArena, a startup that originally launched as a UC Berkeley research project in 2023, announced on Tuesday that it raised a $150 million Series A at a post-money valuation of $1.7 billion. The round was led by Felicis and the university's fund, UC Investments. The startup bolted out of the gate as a commercial venture with a $100 million seed round in May at a $600 million valuation. This new round means it raised $250 million in about seven months. LMArena is best known for its crowdsourced AI model performance leaderboards. Its consumer website lets a user type a prompt that it sends to two models, with the user then choosing which model did a better job. Those results, which now span more than 5 million monthly users across 150 countries and 60 million conversations a month, the company says, fuel the leaderboards. It ranks various models on a variety of tasks including text, web development, vision, text-to-image, and other criteria. The models it tests include various flavors of OpenAI GPT, Google Gemini, Anthropic Claude, and Grok, as well as ones that are geared toward specialties like image generation, text to image, or reasoning. The company began as Chatbot Arena, an open research project built by UC Berkeley researchers Anastasios Angelopoulos and Wei-Lin Chiang, and was originally funded through grants and donations. LMArena's leaderboards became something of an obsession among model makers. When LMArena started pursuing revenue, it partnered with select model companies such as OpenAI, Google, and Anthropic to make their flagship models available for its community to evaluate. In April, a group of competitors published a paper alleging that this helped those model makers game the startup's benchmarks, an allegation LMArena has vehemently denied. In September, it publicly launched a commercial service, AI Evaluations, in which enterprises, model labs, and developers can hire the company to perform model evaluations through its community. This gave LMArena an annualized "consumption rate" - as the company describes its annual recurring revenue (ARR) - of $30 million as of December, less than four months after launch. Join the disrupt 2026 waitlist. Add yourself to the disrupt 2026 waitlist to be first in line when early bird tickets drop. Past disrupts have brought Google cloud, netflix, microsoft, box, phia, a16z, elevenlabs, wayve, hugging face, elad gil, and vinod khosla to the stages - part of 250+ industry leaders driving 200+ sessions built to fuel your growth and sharpen your edge. Plus, meet the hundreds of startups innovating across every sector.
LMArena, open community platform to benchmark and compare ai models through side-by-side evaluations and user voting. Here you'll find information about their funding, investors and team.