> ## Content Index
> Fetch the complete content index at: https://enteruaf.com/llms.txt
> Use this file to discover other available public pages before exploring further.

# a16z wants Silicon Valley to cosign its own reputation graph
- URL: https://enteruaf.com/cosign/
- Published: 2026-09-26T09:24:07.000Z
- Updated: 2026-09-26T09:24:07.000Z
- Description: Cosign turns endorsements and work relationships into a searchable network for people, companies and jobs. Its launch raises a larger question: who gets to curate the professional reputation graph that AI systems will use?
- Author: UAF Editorial
- Tags: Society, Economy

**Cosign turns endorsements and work relationships into a searchable network for people, companies and jobs. Its launch raises a larger question: who gets to curate the professional reputation graph that AI systems will use?**

Hiring has always relied on information that rarely appears on a résumé: who actually worked well with whom, who made a team better, and whose judgment other people trust. Andreessen Horowitz now wants some of that knowledge in public view.

The venture firm’s New Media team launched Cosign on 25 September, describing it as a curated professional network for startup people and companies. Users can endorse people and firms, build lists, answer community questions and follow an Industry Watch feed of company and talent news. The product also connects recommended companies with job openings. [a16z’s launch announcement](https://a16zjobs.substack.com/p/introducing-cosign-a-new-space-to) says its system uses AI to research and populate company data, filter spam and write parts of the feed. The [Cosign homepage](https://cosign.co/?ref=enteruaf.com) presents endorsements as attributable context from people who have worked together.

That pitch targets a real blind spot. A public profile can show a connection; it rarely shows the texture of a working relationship. A former colleague’s specific recommendation can help a person outside a famous company or a well-connected network become visible. A curated list can also tell a job seeker what someone with relevant experience would actually recommend.

But once reputation becomes searchable, the system deciding what to show becomes part of the labor market.

## From private reference to public signal

Cosign’s core unit is a person’s name attached to a judgment. That makes an endorsement useful: it carries some reputational cost for the person giving it. A list of companies one would recommend to a new graduate turns that same gesture into career guidance. Community questions can collect answers that remain attached to profiles, while the Industry Watch feed puts hiring and company activity beside those recommendations.

These features could lower the cost of asking around. They could also reward people whose professional networks are already visible. A platform can display a signal only after someone has supplied it, and absence can be misread as evidence. New entrants, workers with less public-facing jobs, people outside startup hubs and those who do not want their relationships indexed may appear less legible than their actual contributions warrant.

The launch announcement calls the product curated and says AI helps gather information online and filter spam. It does not establish how the system verifies identities, handles inaccurate claims, represents people who have not joined, or lets someone challenge a mistaken profile. Nor does a public endorsement by itself establish job performance. Those details matter because an informal opinion can become a durable input to hiring and investment decisions once it is collected at scale.

## Who curates the map?

Cosign is not an independent industry directory: it is a product built by a16z’s New Media team, which the firm says it plans to expand. That does not invalidate the product. It does make governance and disclosure part of the story. A venture firm’s professional network could be useful to founders and job seekers while also reflecting the people, companies and opportunities most visible to that firm’s ecosystem.

The distinction is easy to miss. “Who is good?” sounds like a social question. “Who is surfaced, by whom, and on what evidence?” is closer to an infrastructure question. If AI systems help create the database and write its news feed, the product’s apparent breadth may partly reflect automated aggregation rather than verified knowledge. Readers should be able to tell those apart.

Cosign may make valuable private knowledge more accessible. The harder test is whether it can preserve context: who is making a recommendation, what experience supports it, what the system inferred, and how a person can correct the record. A reputation graph can make a network easier to search. It cannot make a judgment more reliable merely by giving it a profile page.

The future of hiring may include more public signals about trust. The people who own the index will have a quieter influence over which signals count.