Block, the fintech giant founded by billionaire Jack Dorsey and parent company of Square, Afterpay, and Cash App, has made a bold move in the competitive AI landscape by open-sourcing its AI agent called Goose. Launched on Tuesday, this decision marks a rare strategic pivot in an industry where proprietary technology typically provides competitive advantage.
Goose has demonstrated impressive capabilities since its internal deployment three months ago, automating multi-day engineering tasks in mere minutes and writing code that Brad Axen, Block’s lead AI engineer, describes as “better” and “faster” than some of the company’s top engineers. The AI agent was developed over nine months by a team of approximately 12 specialists, including AI and machine-learning experts, software engineers, and designers, working closely with Anthropic, the Amazon-backed AI startup.
The adoption metrics are compelling: approximately 1,000 engineers (20% of Block’s engineering workforce) have integrated Goose into their workflows, reporting time savings of around 20%. In one striking demonstration, Goose rewrote 70% of a platform’s underlying code in a different programming language in just 30 minutes—a task that would have taken Axen, a principal engineer with a Ph.D. from UC Berkeley, hours or even more than a day to complete.
By making Goose open source, Block is providing any engineer, including competitors, access to its AI agent framework for use, modification, and distribution, even for commercial purposes. Jackie Brosamer, vice president of data and AI engineering at Block, acknowledged this is “a really big bet” for the company, noting that “the risk is that openness isn’t the best way to do a lot of these things.”
Goose’s applications extend beyond engineering. Initially designed to help with writing, editing, and testing code, the tool now assists with automating mundane tasks like migrating legacy systems to new coding languages and implementing updates without disrupting connected systems. Non-technical employees have also begun using Goose for tasks like meeting preparation through Google Calendar integration and writing SQL queries for data analysis. Brosamer stated the company expects Goose to “supercharge by 30% or even more our workforce going forward” without impacting hiring plans.
Key Quotes
People are great at coming up with ideas in a way that AI isn’t yet. Once you agree on the right idea, it writes code better than I, it’s faster than me.
Brad Axen, Block’s lead AI engineer and a principal engineer with a Ph.D. from UC Berkeley, explained how Goose complements rather than replaces human engineers, highlighting the collaborative relationship between AI and human creativity.
Open source is a really big bet that we’re making as a company. The risk is that openness isn’t the best way to do a lot of these things.
Jackie Brosamer, vice president of data and AI engineering at Block, acknowledged the unconventional and risky nature of open-sourcing their AI agent in an industry where intellectual property typically provides competitive advantage.
We’re trying to fulfill this bigger mission of economic empowerment and we see that as more than just beating our competitors or trying to win some section of the market.
Brosamer articulated Block’s strategic rationale for open-sourcing Goose, positioning the decision as aligned with the company’s broader mission rather than traditional competitive dynamics.
We can take those ideas and fold them back to our work rather than getting to it in six months when we needed it.
Brosamer explained how Block expects to benefit from the open-source community’s contributions, accelerating their own development timeline through collaborative innovation.
Our Take
Block’s open-source strategy with Goose represents a fascinating experiment in collaborative AI development that challenges conventional wisdom about competitive moats in technology. While most financial institutions are building proprietary AI agents behind closed doors, Block is betting that the velocity of innovation from community contributions will outpace any first-mover advantage they might sacrifice.
The partnership with Anthropic is particularly noteworthy, as it demonstrates how established companies are leveraging AI startups’ cutting-edge models while building their own application layers. The fact that Goose can work with any large language model provides flexibility and reduces vendor lock-in—a smart architectural decision.
Most intriguing is the expansion beyond engineering use cases to business functions, suggesting AI agents are approaching a tipping point for mainstream enterprise adoption. The 20-30% productivity gains Block reports could become a competitive necessity rather than advantage if these tools become ubiquitous through open-source distribution.
Why This Matters
Block’s decision to open-source Goose represents a significant strategic departure in the intensifying race to develop AI agents on Wall Street and across the financial services industry. While banks, hedge funds, and startups are racing to build proprietary AI assistants for coding, research, and client advisory services, Block is betting that collaborative development will accelerate innovation faster than closed systems.
This move has broader implications for the AI industry’s development model. By sharing its framework built in partnership with Anthropic, Block is potentially accelerating the entire sector’s capabilities while positioning itself to benefit from external contributions and improvements. The company’s philosophy—prioritizing “economic empowerment” over simply “beating competitors”—suggests a vision where widespread AI adoption creates more value than maintaining proprietary advantages.
The productivity gains reported by Block’s engineers (20% time savings, 30% workforce enhancement) provide concrete evidence of AI agents’ transformative potential for knowledge workers. As these tools expand beyond technical roles to business functions, they’re demonstrating that AI agents can augment rather than replace human workers, addressing ongoing concerns about AI’s impact on employment while showcasing how organizations can leverage AI for competitive advantage.
Recommended Reading
For those interested in learning more about artificial intelligence, machine learning, and effective AI communication, here are some excellent resources:
Recommended Reading
Related Stories
- Wall Street Asks Big Tech: Will AI Ever Make Money?
- The Impact of AI on Software Engineering Jobs and Market Outlook
- How Companies Can Use AI to Meet Their Operational and Financial Goals
- Tech Workers Are the Real Winners in the AI Talent War, With Pay Set to Soar by 2024
- Perplexity CEO Predicts AI Will Automate Two White-Collar Roles by 2025
Source: https://www.businessinsider.com/block-afterpay-square-open-source-ai-agent-anthropic-2025-1