Bloomberg’s OpenTelemetry cohort builds trust as AI-generated contributions grow

Bloomberg put 48 engineers into a ten-week OpenTelemetry contribution programme, producing 70 merged pull requests across 11 repositories. The approach offers open-source maintainers something increasingly useful in an AI-shaped flood of code: contributors they know, can mentor and can contact directly.

Watch Desk analysis

What happened

The cohort ran from April to June 2026. Diginomica reports that seven OpenTelemetry maintainers worked with the Bloomberg engineers, most of whom had not contributed to open source before. The programme produced 118 pull requests, of which 70 were merged, and logged 850 volunteer hours.

Bloomberg’s Open Source Program Office arranged the cohort with the Cloud Native Computing Foundation. The $80,000 in philanthropic matching was tied to volunteer hours, with participants able to direct the resulting funds to charities of their choice. It was not a donation straight to OpenTelemetry.

Why it matters

OpenTelemetry maintainer Severin Neumann told Diginomica that the project is receiving more AI-generated contributions, making it harder to judge a contributor’s identity and intent from a pull request alone. He said knowing the Bloomberg participants and having direct channels to them helped maintainers assess their work. The cohort is not an AI detector or a ban on AI-assisted code; contributors and maintainers interviewed for the report said they use AI themselves.

The programme pairs code contributions with relationships and mentorship. That matters for infrastructure projects where maintainers must review outside work, including changes that could affect widely used software.

Our read

This is a practical answer to a trust problem, not a clever filter that promises to spot machine-written code. The useful ingredients are named contributors, ongoing contact and work shaped around maintainers’ priorities. They take time, which rather ruins the fantasy that a fresh tool can automate away the human work.

What to watch

• Whether Bloomberg repeats the cohort with OpenTelemetry or other projects.
• Whether the programme’s model can bring in experienced contributors beyond one large employer.
• How open-source projects adapt review and contributor relationships as AI-assisted submissions become more common. Activity teaser: Bloomberg’s ten-week OpenTelemetry programme paired 48 engineers with seven maintainers, producing 70 merged pull requests. The twist is not a new AI detector: maintainers said knowing contributors and being able to speak to them helped build trust amid a growing flow of AI-generated code. Participants logged 850 volunteer hours, with philanthropic matching going to charities rather than directly to OpenTelemetry. The full story looks at what this model offers open-source projects, and what it cannot promise. Should projects invest more in structured human mentorship, or focus on better technical review tools?

Discussion spark: Should open-source projects invest more in structured human mentorship to build contributor trust, or focus on better technical review tools?

Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.