The Autoscience Institute has introduced ‘Carl,’ the first artificial intelligence (AI) system capable of generating academic research papers that pass a rigorous double-blind peer-review process.

Carl’s research papers were accepted into the Tiny Papers track at the International Conference on Learning Representations (ICLR), with minimal human involvement. This marks a significant milestone in AI-driven scientific discovery, as these submissions were generated with limited human input.

Introducing Carl: The Automated Research Scientist

Carl represents a significant leap forward in the role of AI in academic research, transitioning from a tool to an active participant. Described as an “automated research scientist,” Carl leverages natural language models to ideate, hypothesize, and accurately cite academic work.

Notably, Carl can read and comprehend published papers in a matter of seconds, working continuously to accelerate research cycles and reduce experimental costs. Unlike human researchers, Carl’s tireless operation enables it to process vast amounts of information rapidly.

According to Autoscience, Carl successfully “ideated novel scientific hypotheses, designed and performed experiments, and wrote multiple academic papers that passed peer review at workshops,” demonstrating its potential to complement and even surpass human research in terms of speed and efficiency.

This achievement underlines the potential of AI to not only augment human research but also to expedite the research process, making it more efficient and productive.

Carl’s Work Process: Meticulous and Efficient

Carl’s ability to generate high-quality academic work is built on a three-step process:

Ideation and Hypothesis Formation: Leveraging existing research, Carl identifies potential research directions and generates hypotheses, formulating novel ideas in the field of AI through its deep understanding of related literature.

Experimentation: Carl writes code, tests hypotheses, and visualizes the resulting data through detailed figures, shortening iteration times and reducing redundant tasks through its tireless operation.

Presentation: Finally, Carl compiles its findings into polished academic papers, complete with data visualizations and clearly articulated conclusions.

Although Carl’s capabilities make it largely independent, there are points in its workflow where human involvement is still necessary to ensure adherence to computational, formatting, and ethical standards:

• Greenlighting Research Steps: Human reviewers provide “continue” or “stop” signals during specific stages of Carl’s process, guiding it through projects more efficiently without influencing the research itself.
• Citations and Formatting: The Autoscience team ensures all references are correctly cited and formatted to meet academic standards, currently a manual step but essential for aligning the research with publication venue expectations.
• Assistance with Pre-API Models: Carl occasionally relies on newer OpenAI and Deep Research models lacking auto-accessible APIs, requiring manual interventions like copy-pasting outputs to bridge these gaps, which Autoscience expects to automate in the future when APIs become available.
For Carl’s debut paper, the human team also assisted in crafting the “related works” section and refining the language, tasks that were unnecessary following updates applied before subsequent submissions.

Rigorous Verification Process for Academic Integrity

Before submitting any research, the Autoscience team undertook a stringent verification process to ensure Carl’s work met the highest standards of academic integrity:

• Reproducibility: Every line of Carl’s code was reviewed, and experiments were rerun to confirm reproducibility, ensuring the findings were scientifically valid and not coincidental anomalies.
• Originality Checks: Autoscience conducted extensive novelty evaluations to ensure Carl’s ideas were new contributions to the field and not rehashed versions of existing publications.
• External Validation: A hackathon involving researchers from prominent academic institutions, such as MIT, Stanford University, and U.C. Berkeley, independently verified Carl’s research, with further plagiarism and citation checks performed to ensure compliance with academic norms.

Undeniable Potential and Larger Questions

Achieving acceptance at a respected workshop like the ICLR is a significant milestone, but Autoscience recognizes the broader conversation this milestone may spark. Carl’s success raises larger philosophical and logistical questions about the role of AI in academic settings.“We believe that legitimate results should be added to the public knowledge base, regardless of their origin,” explained Autoscience. “If research meets the scientific standards set by the academic community, then who – or what – created it should not lead to automatic disqualification.”“However, we also believe that proper attribution is necessary for transparent science, and work purely generated by AI systems should be discernible from that produced by humans.”Given the novelty of autonomous AI researchers like Carl, conference organizers may need time to establish new guidelines that account for this emerging paradigm, ensuring fair evaluation and intellectual attribution standards. To prevent unnecessary controversy, Autoscience has withdrawn Carl’s papers from ICLR workshops while these frameworks are being devised.Moving forward, Autoscience aims to contribute to shaping these evolving standards, intending to propose a dedicated workshop at NeurIPS 2025 to formally accommodate research submissions from autonomous research systems.As the narrative surrounding AI-generated research unfolds, it’s clear that systems like Carl are not merely tools but collaborators in the pursuit of knowledge. However, as these systems transcend typical boundaries, the academic community must adapt to fully embrace this new paradigm while safeguarding integrity, transparency, and proper attribution.(Photo by Rohit Tandon)

See also: You.com ARI: Professional-grade AI research agent for businesses



Want to learn more about AI and big data from industry leaders? Check out the AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events, including the Intelligent Automation Conference, BlockX, Digital Transformation Week, and the Cyber Security & Cloud Expo.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.