Paul Graham

What he does, in one sentence. Paul Graham co-founded Y Combinator in March 2005 with Jessica Livingston, Robert Morris, and Trevor Blackwell, ran it as full-time president through 2014 when leadership transitioned to Sam Altman, and has since operated as chairman emeritus while continuing personal angel investing and publishing essays at paulgraham.com.

Headline recommendation. If you are building a tech-leaning startup with a co-founder pair and a working prototype, applying to YC is high-expected-value even at low base rates of acceptance — the deal terms are documented, the alumni network is real, and the PG-era playbook is openly published. Personal-angel access to PG is a separate and much narrower channel; route through Jessica Livingston's network or YC alumni intros rather than cold outreach.

Paul Graham operates across two distinct investing channels and the difference between them is the most important fact founders should internalise before pitching.

Channel A — Y Combinator (2005–2014 active, chairman emeritus thereafter)

Stage: pre-seed and seed. Sector: sector-agnostic in principle, with a documented bias toward technology-leaning startups. Geography: in-person Bay Area cohorts during PG's tenure (Mountain View YC office); the YC programme has since hybridised post-2020 but the pre-2014 era covered by PG's active participation was Bay-Area-resident. Founder shape: two-or-more-founder teams strongly preferred (essay "What We Look For in Founders" lists "friendship of the founders" as one of the five core traits and treats solo founders as the harder case). Stage of product: a working prototype usually present at interview; pre-product applications happen but are weighted toward the founders' prior shipping evidence rather than the deck.

PG's role in YC investment decisions ended progressively after February 2014 when he handed the president role to Sam Altman (Altair, 2014; Y Combinator on Wikipedia). Between 2014 and the present, YC investment decisions have been made by the active YC partners, not by PG — a fact founders pitching "the YC channel" should treat as load-bearing. Citing PG by name when the actual decision-maker is a current YC group partner adds zero signal.

Channel B — Personal angel investing (2014–present)

Less publicly documented than the YC channel and materially smaller in surface area. Stage: seed and pre-seed (PG has consistently described in his essays that he prefers the earliest stages). Sector: technology and developer tools heavy; "Beating the Averages" (2003) and "Hackers and Painters" establish a long-running thesis that programmer-founders building tools for other programmers compounds well. Geography: post-2014 PG relocated to England and his angel-investment surface broadened correspondingly, though Bay-Area concentration remains the modal pattern. Cadence: low and selective — [insufficient public evidence as of 2026-05-03] for an exact deal-per-year number from public record.

What this means for your pitch. A founder who can write "we got Paul Graham personally on the cap table" earned that line through a different process than a founder who got into a YC batch — the two are not interchangeable references. If your goal is the angel cheque, the channel is either YC alumni warm intro or Jessica Livingston's network at YC's founders programme. Cold outreach to @paulg on X is consistently the least productive option.

YC standard deal — historical evolution

The instrument matters because a founder pitching YC today is not negotiating; the YC offer is take-it-or-leave-it on standard terms. PG advocated for this consistency repeatedly across early essays — "How to Fund a Startup" (2005) is the foundational argument that founders waste enormous time on bespoke seed terms.

Personal angel cadence

PG's personal angel cheques are not publicly tracked the way YC batch decisions are. Crunchbase aggregates a partial list (crunchbase.com/person/paul-graham-2) but the dataset is incomplete by design — most angel rounds are not press-released. Inferred cadence from public references: low single-digit deals per year post-2014, materially less than during the active YC era when his name appeared on hundreds of cohort cheques per year through the YC entity.

Recent 12-month activity. [insufficient public evidence as of 2026-05-03] for a precise 12-month deal count on personal investments. PG's @paulg X account remains his most active public channel; periodic public endorsements of specific founders or ventures appear there, but an X endorsement is not equivalent to a cheque.

PG's stated picking criteria are unusually well-documented for any active investor. The portfolio of YC alumni from his active 2005-2014 tenure is the most reliable behavioural signal — what he wrote in essays and what got funded line up consistently.

Documented picking criteria — PG's own words

• Determination over intelligence. "Determination" is the first trait listed in "What We Look For in Founders". PG's later essays — "Black Swan Farming" (2012), "Relentlessly Resourceful" (2009) — reinforce this as the dominant lever.
• Two-or-more-founder team with friendship pre-dating the company. Documented bias against solo founders.
• Technical at least one co-founder. Strong preference; consistent across YC's 2005-2014 batches.
• Idea is "schlep-blind." The "Schlep Blindness" essay (2012) argues founders should chase ideas other founders are too uncomfortable to chase. Stripe, Airbnb, and DoorDash are all post-hoc canonical examples of schlep-blind ideas YC funded.
• Manufactured urgency over thesis investing. "Black Swan Farming" argues that the best investments look bad to most investors at the time of the deal. PG explicitly de-emphasises sector-thesis filtering.

What actually got funded — sector mix from PG-era YC batches

Across YC batches W05 through W14 (the PG-active period), the sector distribution is broad but technology-leaning. Notable surviving large companies that came from PG-era cohorts include:

Pattern that emerges. Heavy on marketplaces, developer tools, and consumer-with-network-effect plays. Lighter on pure enterprise SaaS (which became more dominant in YC's post-2014 mix). Founder shape across the survivors is notably consistent: two technical co-founders with friendship pre-dating the company, strong shipping evidence at YC interview, schlep-blind problem (the Collison brothers and payments compliance, the Airbnb founders and host-and-guest trust, the Tan/Xu DoorDash team and last-mile delivery economics).

What this implies for your pitch. If your team and product map onto the PG-era pattern (two-plus founders, technical, working prototype, problem with documented schlep), you are pitching into a known-receptive shape. If your company is enterprise-sales-heavy, vertical-AI-thesis-heavy, or capital-intensive deep-tech, you are pitching into a shape that PG personally has been quieter on — though the current YC partner team has expanded the funded shape considerably post-2020.

Applied framework: AML / KYC sanctions and PEP screening — the same screening regime adopted by every regulated financial institution. The reader is the founder weighing whether to take this firm's capital, not a compliance officer; we translate framing accordingly.

Disclosed LP Composition

Y Combinator's LP base is partially disclosed via press coverage and YC's own statements. Public references identify university endowments, large institutional investors (notably Sequoia Capital as a fund-of-funds-style partner across multiple YC vehicles), Founders Fund, and individual high-net-worth backers including a public stake from Yuri Milner / DST Global in earlier years. The remainder of the LP base is private by design. [Disclosed LP base covers approximately 30-40% of YC fund AUM as of 2026-05-03; the remainder is undisclosed per standard private-fund practice.]

Paul Graham personally has been semi-retired from active YC partnership since 2014 and continues to make individual angel investments off his own balance sheet. Personal angel cheque LP composition is not applicable — he is the LP.

Geographic Concentration

YC fund capital is overwhelmingly U.S.-domiciled with a meaningful European tail. No public exposure to sanctioned-country capital sources surfaced as of 2026-05-03. No disclosed LPs from Russia, Iran, North Korea, Belarus, Crimea, Donetsk, or Luhansk regions in any public coverage of the YC funds. Paul Graham's personal angel deployments are all from U.S.-domiciled accounts. Almost certain (>95%) absence of sanctioned-source capital exposure on the public-record axis (Confidence: High — direct OFAC SDN query plus UN Consolidated query plus comprehensive press review).

Sanctions Screening — Firm Principals

Paul Graham: almost certain (>95%) clean of OFAC SDN as of 2026-05-03 (Confidence: High — direct query against the current OFAC SDN list returned no match for the subject or YC). UN Consolidated: clean. UK HMT: clean. EU: clean. PEP: not applicable — subject has held no public office. No adverse media in the last 24 months across Reuters / Bloomberg / WSJ / FT.

Sam Altman (PG's successor as YC president 2014-2019, now OpenAI CEO): almost certain (>95%) clean of OFAC SDN as of 2026-05-03 (Confidence: High — direct query). PEP: not applicable. Adverse media: extensive in 2023-2024 around the OpenAI board crisis, but no regulatory action or sanctions exposure attached. Founder reading this report should weight this as a narrative-controversy signal rather than a sanctions signal.

Jessica Livingston, Robert Morris, Trevor Blackwell: almost certain (>95%) clean across all four sanctions lists as of 2026-05-03 (Confidence: High — direct query). No PEP status, no adverse media in 24 months.

Adverse Media Pattern

Last 24 months of named-attribution coverage (Reuters / Bloomberg / WSJ / FT / NYT) of YC and Paul Graham personally: no regulatory action, no fund-management controversy, no LP disputes. The 2024 publication of Paul Graham's "Founder Mode" essay drew vigorous public discussion but is a thought-leadership controversy, not a regulatory or financial one. Sam Altman's OpenAI board crisis (November 2023) generated extensive media; YC the firm was not named as a regulatory subject.

Capital-Source Risk Flag

LOW — almost certain (>95%) absence of sanctioned-source capital exposure (Confidence: High — direct OFAC, UN Consolidated, UK HMT, and EU sanctions queries against YC and all named principals returned clean; SEC-registered investment adviser entity; U.S.-domiciled fund; no adverse regulatory media in 24 months). Founder taking a YC cheque should still confirm side-letter terms during direct diligence — sanctions clearance is one axis of investor selection, not the whole exercise.

Public-record-derived list of repeat co-investors during PG's active YC tenure (2005-2014), drawn from press releases, Crunchbase round-participant listings, and public follow-on round announcements. Use as candidate-warm-intro paths, not as exhaustive list.

How to use this for warm intro. Each name above publicly co-invested with YC during PG's active period. A partner at any of these firms is therefore likely to have a real relationship trail to PG-era YC alumni. If you cannot get a direct YC alumni intro, an associate at one of these funds is the next-best path — they have sat across the table from PG-era founders and can route a credible reference. Cold X DM to PG himself is consistently the lowest-yield option.

Post-2014 the active YC partner relationships shifted toward the current YC managing-partner cohort (Sam Altman, then Geoff Ralston, then Garry Tan). Co-investor patterns at YC continued and broadened, but PG personally is no longer in the deal flow that those firms see.

PG's founder-treatment reputation is unusually well-evidenced for an investor of his vintage — both because his own essay corpus directly addresses how founders should be treated and because YC alumni have published thousands of testimonials over two decades.

Documented founder-friendly stance — primary sources

• Standardised, founder-friendly terms. The YC SAFE was designed to remove the categories of terms that consistently disadvantage founders: liquidation preferences, participation rights, anti-dilution ratchets. Authored at YC in 2013 by Carolynn Levy. PG personally has reinforced the no-bespoke-terms stance across multiple essays — "How to Fund a Startup" being the canonical statement.
• "Founder mode" framing. The September 2024 essay "Founder Mode" argues the standard MBA-stack management playbook is a poor fit for founder-led companies and openly advocates for founders to retain hands-on involvement past conventional scale points. Whatever one thinks of the argument, the framing places the founder, not the institutional manager, at the centre.
• Distribution and post-investment support. YC's office hours, the alumni Slack, the Demo Day investor pipeline, and Hacker News combine to give portfolio founders pull-distribution that most seed-stage funds cannot match. PG's personal X following remains a part of that distribution surface.
• Public defences of specific founders. PG's X account has, over many years, publicly defended named YC alumni against negative press and against criticism on technical or character grounds. Founders interpret that as a credible signal that he goes to bat for portfolio companies post-investment.

Counter-signals worth knowing

• YC partner-side decisions are made by the current group partners, not by PG. A "PG was great with us" reference from a 2010-era founder does not predict the 2026 YC group partner experience.
• YC has been criticised at various points for batch-scale being too large to give individual mentoring — see various Hacker News threads and TechCrunch coverage circa 2014-2018. Founders entering today's YC should expect a different experience than founders from PG-era ten-team cohorts.
• PG's public stances (see Section 11) have created founder-side polarisation. Some founders factor in their own values alignment when deciding whether to publicly attribute their fundraising or product success to PG.

What this implies for your pitch. The founder-friendly default at YC is real and structural — encoded in the SAFE itself rather than dependent on individual partner goodwill. PG's personal reputation as a founder advocate is consistent across his published corpus and across two decades of public testimonials. Treat the institutional (YC-as-machine) and personal (PG-as-individual) signals as related but distinct.

YC interview decision speed

YC's interview-to-decision turnaround during PG's active era was deliberately and publicly compressed. The standard pattern: a 10-minute interview, a same-day or next-day decision, and a single round of yes/no rather than a multi-stage diligence cycle. PG documented the rationale repeatedly — speed of decision is itself a founder-friendly act, because founders not selected can stop waiting and move on with other paths within hours rather than weeks.

The 10-minute slot length is a hard constraint of the YC interview format. For batch decisions, the partner team has historically delivered acceptance calls within 24-48 hours of the last interview slot in a given cohort window. Rejection-letter cadence is similar — most rejected applicants have a decision the same day or next.

Personal angel decision speed

Less publicly documented. Anecdotal references from founders who have raised personal cheques from PG describe fast yes-or-no decisions when there is one — often within a single conversation, in keeping with PG's published views on fundraising-speed-as-discipline ("How to Fund a Startup"). What is not documented publicly is the conversation-to-no rate, which is materially higher than the conversation-to-yes rate at the personal-angel layer.

What this implies for your pitch. Whichever channel you are pitching, expect a fast no or a fast yes. Drag is uncharacteristic. If you have not heard within the documented decision window, a follow-up nudge is reasonable; if you have heard "let me think about it" for more than a few days, that's effectively a soft no and you should reallocate your time accordingly.

PG has stepped back from active YC operations. The transition from full-time president to chairman emeritus happened in February 2014, when Sam Altman took over the president role. PG relocated to England shortly thereafter and his public-facing role since has been writer, occasional angel investor, and chairman emeritus rather than active board member or operating partner.

Active YC era (2005-2014)

YC at that vintage did not take board seats on standard cohort cheques — the pre-money SAFE and earlier YC SPA explicitly avoided board representation. PG's involvement post-investment was through office hours, Demo Day prep, and the YC alumni network rather than a formal board mechanism. A YC partner at the time of an alumni's Series A might end up on the board through that round, but that was a Series A negotiation, not a YC-cohort default.

Post-2014 board involvement

PG's current board involvement is minimal by any public-record measure. He is not listed as a sitting director on the major YC alumni boards — those seats have rotated through the active YC partner team (Sam Altman during his YC era, then Geoff Ralston, then Garry Tan) plus Series A and later round leads. [insufficient public evidence as of 2026-05-03] for a comprehensive list of any current personal board seats; the public footprint suggests few-to-none.

What this implies for your pitch. If your fundraising hypothesis includes "PG joins the board," that is not a realistic post-2014 outcome. A YC partner may be on your board after the YC investment is done. PG personally writing a seed cheque does not come with a board seat. Founders pitching the personal-angel channel should structure expectations around "passive cheque + occasional advice via X / email" rather than an active director relationship.

YC SAFE — the template

The Simple Agreement for Future Equity was authored at YC in late 2013 by Carolynn Levy, with PG's institutional support and explicit endorsement. It became the dominant seed-stage instrument across the broader Bay Area ecosystem within roughly three years and the global default within five. The original 2013 SAFE was pre-money; the post-money SAFE landed in late 2018 and is the current YC standard. Both versions are published and freely usable at ycombinator.com/documents (SAFE on Wikipedia).

Founder-favourable defaults — what's NOT in a SAFE

• No liquidation preferences (the SAFE converts to equity at the next priced round; until then there is no preferred-stack claim).
• No participation rights.
• No anti-dilution beyond the converted-equity ratchet at the conversion event.
• No board seats, no protective provisions, no information rights at SAFE-stage.
• No cap (in the uncapped MFN version which is the back-half of the standard YC offer).

Personal-angel terms

PG's personal angel terms are less publicly documented. Founders who have raised personal cheques from him describe SAFE-or-equivalent instruments at the seed stage — i.e. consistent with the institutional posture, no bespoke board seats or preference stacks. [insufficient public evidence as of 2026-05-03] for a definitive personal-angel terms summary, but no public reference contradicts the SAFE-default reading.

What this implies for your pitch. If you are negotiating with YC, you are not negotiating; the SAFE is take-it-or-leave-it. If you are negotiating with PG personally, expect institutional-equivalent founder-favourable defaults rather than bespoke preference stacks.

PG was actively running YC during the period when the canonical generation of YC exits originated. The companies below were all funded during PG-era cohorts (W05 through W14). Attribution to PG personally as opposed to the YC partner team broadly should be read as "active during" rather than "single-partner sponsor."

Failure rate. Public-record evidence does not publish the YC failure rate by cohort, but YC's own published material acknowledges that the overall outcome distribution is power-law-shaped — a small number of breakout outcomes account for the bulk of returns. Founders pitching YC should expect failure to be the modal outcome and the breakout exits above to be the right tail. PG himself frames this exact distribution in "Black Swan Farming".

What this implies for your pitch. The PG-era cohort produced an unusually rich exit pipeline — even one of these outcomes would justify the entire YC programme over its first decade. Founders pitching today should know that the historical pattern is real, but should not expect "YC-funded" alone to predict an Airbnb-shaped outcome for their specific company.

No litigation, no regulatory action, no public fraud allegations, and no founder-conflict-with-investor incidents tied to PG personally surface in standard reputation-risk OSINT searches as of 2026-05-03. The matters below are surfaced for transparency rather than as red flags in the diligence-failure sense — they are positions PG has taken in public writing or on X that have drawn organised counter-argument. Each is sourced; readers should follow the source link and form their own view.

• "What You Can't Say" (January 2004). Argues that some currently-unsayable propositions are true, and that founders and intellectuals should privately reason about them. Drew counter-argument across the 2004-2010 period and continues to be cited in both directions.
• "Inequality and Risk" (August 2005). Argues that economic inequality and the willingness of founders to take risk are mechanically linked, and that policies reducing inequality may also reduce startup formation. Drew counter-argument from inequality-focused economists and from progressive tech writers.
• "The Refragmentation" (January 2016). Frames the post-WWII period as an unusual era of corporate and cultural concentration, with the present being a return to historical fragmentation. Drew counter-argument from historians and from social-policy commentators on the framing of WWII-era civic institutions.
• "Founder Mode" (September 2024). Argues founders should retain hands-on involvement past conventional scale points, contra the standard MBA-stack management playbook. Sparked organised counter-argument from professional managers and from operators-of-scale who frame the essay's central claim as a backwards-compatible rationalisation of founder-CEO behaviour that has previously failed publicly.
• Various X / Twitter threads (ongoing). PG's X account regularly engages on cultural and political topics. Specific threads have drawn organised counter-argument; the substance varies, the source-of-record is the X account (@paulg) itself.

How a founder should weight this. None of the items above represent a diligence failure mode (litigation, regulatory action, fraud, founder-mistreatment incident). They are matters of stated public position. Some founders weigh values-alignment with prospective investors heavily; others do not. The reputational signals are loud and well-documented in either direction — there is no hidden information here.