Who Is Daniel Kokotajlo, Lead Author of AI 2027?

Daniel Kokotajlo is an artificial intelligence (AI) researcher, the executive director of the AI Futures Project, and the lead author of the AI 2027 forecast scenario.

At Layer3Labs, we build AI agents for small-business workflows, so we read timeline forecasts like his against what agents can already do in client work.

Kokotajlo previously worked on scenario planning in the governance division of OpenAI before resigning in April 2024, saying he had lost trust in OpenAI leadership.

Daniel Kokotajlo in Brief

Kokotajlo is Executive Director of the AI Futures Project, a forecasting nonprofit based in Berkeley, California. On September 30, 2026, he testified to a Senate subcommittee about when AI research itself could be automated, as documented in his Senate testimony transcript.

He left OpenAI in April 2024 and declined to sign its exit non-disparagement agreement, which put a large sum of his equity at risk, according to The Decoder. His team now publishes regular updates on when AI could fully automate coding and AI research.

• Current role: Executive Director of the AI Futures Project.
• Organization: AI Futures Project, a Berkeley-based 501(c)(3) nonprofit formerly known as Artificial Intelligence Forecasting Inc.
• Known for: Serving as lead author of the AI 2027 scenario, co-authoring AI 2040: Plan A, and signing the Right to Warn open letter.
• Education: Bachelor of Arts (BA) in philosophy from the University of Notre Dame (2014); Master of Arts (MA) in philosophy from the University of North Carolina at Chapel Hill (2017).
• In the news: Testifying before the United States Senate in September 2026 and appearing on CNBC's Squawk on the Street.

Realism and Trajectory of the AI 2027 Model

AI 2027 depicted fully autonomous AI agents better than humans at "everything" around the end of 2027, and the authors' own tracking since then shows real progress running slower than the scenario. In April 2025, the AI Futures Project published AI 2027, authored by Daniel Kokotajlo, Eli Lifland, Thomas Larsen, and Romeo Dean, as outlined on the AI 2027 project page. The Guardian reported that the team named 2027 as its "most likely" guess for "fully autonomous coding."

Subsequent revisions by the research team extended those dates. On December 31, 2025, the AI Futures Model update pushed the median projection for full coding automation out by about 3 to 5 years. Reporting by The Guardian noted that this update shifted autonomous programming projections into the early 2030s and superintelligence horizons to 2034. Kokotajlo wrote on X, as quoted by The Guardian: "Our timelines were longer than 2027 when we published and now they are a bit longer still."

In February 2026, the AI Futures Project published Grading AI 2027's 2025 Predictions, which found progress on quantitative metrics running at roughly 65% of the pace AI 2027 predicted, a figure later revised to about 75% as of July 2026. The group's Q1 2026 update moved Kokotajlo's Automated Coder median from late 2029 to mid-2028. The Q2.5 update in August 2026 said timelines had become slightly shorter and that reality seems to be going about 70% to 90% as fast as AI 2027 predicted.

OpenAI Governance Research and Resignation

Kokotajlo was a researcher in OpenAI's governance division from 2022 to 2024. The AI Futures Project about page describes that work as scenario planning.

In statements reported by Vox and quoted by The Decoder, Kokotajlo explained his departure: "I gradually lost trust in OpenAI leadership and their ability to responsibly handle AGI, so I quit." In his September 2026 Senate testimony, he added: "I resigned partly due to losing confidence that the company would behave responsibly, and partly so I could speak more freely."

In June 2024, he joined other former and active lab researchers in publishing the Right to Warn open letter.

Structure and Funding of the AI Futures Project

The AI Futures Project was founded after Kokotajlo left OpenAI, per Wikipedia's AI Futures Project entry, and its own about page says it was previously named Artificial Intelligence Forecasting Inc.

The AI Futures Project says its funding comes primarily from the Survival and Flourishing Fund alongside a major individual donor. The research staff evaluates benchmark trajectories, creates quantitative models, and publishes scenario updates on their research archive.

Policy Framework in AI 2040: Plan A

On July 9, 2026, the AI Futures Project released AI 2040: Plan A, outlining strategic recommendations for global artificial intelligence governance. The proposal was led by Thomas Larsen and co-authored by Daniel Kokotajlo, Eli Lifland, Romeo Dean, Ryan Greenblatt, and Brendan Halstead.

The authors write: "It's called Plan A because it's a recommendation, not a prediction." In the Plan A scenario, action by the US and Chinese governments delays the creation of superintelligence to 2040; the authors say it would otherwise have happened in 2030.

According to Semafor, Plan A advocates for the US and China to cooperate on a temporary pause of frontier AI research. "We're not exactly de-growthers," Kokotajlo told Semafor.

Senate Testimony and Regulatory Proposals

On September 30, 2026, Kokotajlo testified before the Senate Subcommittee on Disaster Management, District of Columbia, and Census, chaired by Senator Josh Hawley.

In his testimony, Kokotajlo gave his own estimate for when AI research itself could be automated: "I personally would guess about 50% chance by end of 2028."

Kokotajlo outlined two concrete policy priorities for legislators: dramatically improving transparency into the AI industry, and redirecting compute away from automating AI research. Earlier that month, on September 11, 2026, he appeared on CNBC's Squawk on the Street.

Academic Training and Scenario History

Kokotajlo studied philosophy before working on AI forecasting. According to Wikipedia's biographical profile, he graduated with a Bachelor of Arts in philosophy from the University of Notre Dame in 2014 and received a Master of Arts in philosophy from the University of North Carolina at Chapel Hill in 2017.

His scenario modeling work predates his arrival at OpenAI. In 2021, Kokotajlo published "What 2026 Looks Like," an AI scenario forecast for 2022 to 2026. TIME named him to the TIME100 AI list in 2024.

Contrasting Approaches to Artificial Intelligence Safety

Kokotajlo forecasts from outside the labs and grades AI 2027 against measured progress. Anthropic co-founder Chris Olah works from the inside, reverse engineering neural networks into human-understandable algorithms.

Sholto Douglas works inside Anthropic on scaling reinforcement learning and has said he believes AGI is likely within the next couple of years. More people in these debates appear in our summary of AI people to watch.