99 research idea notes tagged “ai”. Read them as Markdown, JSON or an Atom feed.
Text-to-SQL, table question answering and semantic parsing appear on only 12 recent papers, all in AI journals, none in business ones. Firms are rolling out assistants that turn managers' questions into database queries. How often do the queries return the wrong numbers, and would a manager notice?
Psychological safety and employee voice appear on 21 recent papers, almost all in management journals, and only 2 of them carry an AI keyword. As AI tools enter everyday work, someone has to notice and report when they are wrong. Do teams with psychological safety catch AI errors that others let pass?
Papers in the last 30 days — employee voice: 1
Process mining and robotic process automation appear on only 25 recent papers, mostly in AI and operations research, with 3 in accounting. Event logs from enterprise systems record every step of purchasing, invoicing and payment. Can auditors test whole populations of transactions instead of samples, and what do they find?
Non-GAAP earnings appear on only 6 recent papers. Technology firms report profits that exclude stock-based pay and, increasingly, the costs of AI investment, while AI start-ups report "adjusted" margins of their own design. Which adjustments mislead, and do investors see through them?
Qualitative methods appear on 46 recent papers, mostly in management, and only 7 of them carry an AI or language-model keyword. Researchers are starting to let language models code interviews and open-ended answers. Do the codes agree with trained human coders, and what does the method section need to report?
Papers in the last 30 days — fuzzy-set qualitative comparative analysis: 2
Emotion recognition and affective computing appear on 325 recent papers, almost all in AI journals; only 19 carry a consumer, marketing or ethics keyword. The EU now bans emotion recognition at work and in schools, but not in shops or advertising. Does reading faces and voices tell marketers anything reliable, and should it be allowed?
Papers in the last 30 days — emotion recognition: 2 · affective computing: 2
Annual reports, 10-K filings, narrative disclosure and readability appear on 99 recent papers in the finance and accounting journals, but only 9 of them carry a generative-AI, language-model or AI-generated keyword. If language models now draft the management discussion, do the text measures researchers rely on still mean anything?
Papers in the last 30 days — narrative disclosure: 2 · readability: 1
Pseudo-labeling, label noise and weakly supervised learning appear on nearly 500 recent papers, almost all in AI journals, and only 7 of them concern financial, legal or business documents. Banks and auditors have millions of unlabelled contracts, invoices and filings and few labelled ones. Can weak supervision from rules and models replace hand labelling, and what errors does it introduce?
Papers in the last 30 days — weakly supervised learning: 10 · label noise: 4
Earnings calls are a staple of finance and accounting research, read almost always as transcripts. In the AI journals, multimodal fusion is one of the biggest topics. Do voice and video carry information about firms that the words miss?
Papers in the last 30 days — earnings conference calls: 5 · earnings call transcripts: 1
Virtual influencers, avatars and virtual humans appear on 33 recent papers in the management and AI journals, with 9 in the last quarter against about 7 per quarter before. Brands now run computer-generated personalities with millions of followers. Do they sell, do followers care that no one is there, and what should disclosure rules require?
Papers in the last 30 days — virtual influencers: 1
University spin-offs and university entrepreneurship are among the fastest-growing keywords in the management journals, while AI research shifts from universities to companies. Only 2 of 37 recent papers on academic entrepreneurship mention AI. Do universities still turn AI research into startups, or does industry absorb the talent first?
Papers in the last 30 days — university spin-offs: 1 · technology commercialization: 1
Transfer learning, domain adaptation and domain generalization appear on nearly 700 recent papers, almost all in AI journals, and only 10 of them concern credit, lending, banking or markets. Lenders entering new countries, and credit bureaus in countries with little data, want to reuse models trained elsewhere. How much transfers, and what breaks?
Papers in the last 30 days — transfer learning: 9 · domain adaptation: 14
Clinical decision support appears on 94 recent papers and medical image segmentation on 188, among more than 1,300 medical AI papers in the AI journals, but only 65 of all medical papers carry a cost, reimbursement, deployment or workflow keyword, and cost-effectiveness appears on 27 papers in the whole archive. Which medical AI tools pay for themselves, and for whom?
Papers in the last 30 days — clinical decision support: 3 · cost-effectiveness: 1
Low-resource languages are a fast-growing topic in the AI journals, and the digital divide is rising in economics. Language models work worse in languages with little training text, which are spoken mostly in poorer countries. Does that slow AI adoption, and its economic gains, where those languages are spoken?
Papers in the last 30 days — low-resource languages: 1 · digital divide: 9
Inference latency and resource-constrained deployment are among the fastest-growing keywords in the AI journals, yet only 15 of 226 such papers carry an energy, carbon or green-AI keyword. How much computing do the methods in AI journals cost to train and run, and would reporting it change which methods win?
Papers in the last 30 days — inference latency: 4 · resource-constrained deployment: 2
Tabular data appears on 53 recent papers and tabular classification on 13, almost all in AI journals, and 16 tabular papers involve pretrained or language models. Most business data is tables: loans, transactions, patients, sales. Do pretrained tabular models beat gradient boosting on real business tables, and when?
Papers in the last 30 days — tabular data: 4 · tabular data classification: 1
Diffusion models are among the largest topics in the AI journals, and synthetic data generation is growing with them. Yet only 6 of 119 recent synthetic-data papers carry a finance keyword. Could synthetic transaction and loan data let banks share data for research without exposing customers, and how close is it to the real thing?
Papers in the last 30 days — synthetic data generation: 3 · diffusion models: 18
Mamba and other selective state-space models are among the most frequent new architectures in the AI journals, built for long sequences. Limit order books produce exactly such sequences. Yet only 4 of 329 recent Mamba and state-space papers carry a finance keyword, all of them in AI journals. Can these models forecast short-term price moves better, and is the gain tradable after costs?
Papers in the last 30 days — Mamba: 10 · limit order book: 1
Knowledge distillation, model compression and small language models appear on more than 700 recent papers in AI journals, and only 12 of them concern finance, banking, legal or business tasks. Banks cannot send customer data to the largest models, and regulators want models they can inspect. How small can a model be and still do financial work, and what is lost in the shrinking?
Papers in the last 30 days — knowledge distillation: 23 · small language models: 1
Robo-advisors appear on 12 recent papers, and the topic is fading: 1 paper in July–September 2026 against about 5 per quarter before. Meanwhile people ask chatbots for financial advice. Does conversational advice change portfolios more than the old questionnaire-based robo-advice did, and for better or worse?
Papers in the last 30 days — robo-advisors: 1 · household portfolio choice: 2
Wind power, photovoltaic power and solar irradiance forecasting appear on 52 recent papers, almost all in AI journals, and only 8 of them carry a market, bidding, price, imbalance or trading keyword. Forecasts are judged by error metrics, but wind and solar operators pay for forecast errors in imbalance charges. Which forecast improvements pay, and which do not?
Papers in the last 30 days — photovoltaic power forecasting: 1
Recommender systems appear on 111 recent papers and personalized recommendation on 54, almost all in AI journals; 45 recommendation papers carry a diversity, fairness or popularity-bias keyword, and welfare, consumer surplus and addiction appear in almost none. Recommendation decides what billions of people see and buy. Who gains, and by how much?
Papers in the last 30 days — recommender systems: 9
Document understanding, table question answering and chart understanding appear on 17 recent papers, all in AI journals, with 7 in the last quarter against about 3 per quarter before. Financial information lives in tables, footnotes and charts that models still misread. How accurate are current models on real filings, and which errors would mislead an analyst?
Retrieval-augmented generation appears on 165 recent papers, 171 of 178 such papers in AI journals, and only 12 touch finance, legal, tax, accounting, audit or compliance questions. Banks, tax advisers and auditors are deploying it over filings, rules and contracts. How often does it cite the right source, and what happens when the source changes?
Papers in the last 30 days — retrieval-augmented generation: 14
Quantum computing, quantum machine learning and quantum annealing appear on 115 recent papers, mostly in AI journals, and 33 of them concern optimization. Firms in finance and logistics run quantum pilots. Do the quantum results beat the best classical solvers on the same instances, or only the classical baselines the papers chose?
Papers in the last 30 days — quantum computing: 1 · quantum machine learning: 2
Prompt injection, jailbreak attacks and LLM security appear on 37 recent papers, all in AI journals. Banks and brokers are deploying agents that read emails, documents and web pages and then act. A hidden instruction in a document can make such an agent move money or leak data. How exposed are deployed financial agents, and which defences hold?
Papers in the last 30 days — prompt injection: 1
Precision and smart agriculture, plant disease detection and agricultural technology adoption appear on 148 recent papers, split between AI and economics; 47 of them carry an adoption, farmer, cost or income keyword. AI papers detect disease on leaves; economics papers study adoption of older technologies. Do the new tools reach farms, and do they raise income?
Papers in the last 30 days — precision agriculture: 1 · smart agriculture: 1
Open-source software appears on 13 recent papers and open-source language models on 6, while named open-weight models such as DeepSeek, Llama and Qwen appear on dozens of AI papers as tools; 78 of 126 open-source papers involve language models. Firms give away models worth hundreds of millions to train. Who adopts them, and what do the givers get?
Open innovation and crowdsourcing appear on 71 recent papers, mostly in management, but only 3 of them carry an AI keyword. Firms ran idea contests and crowdsourcing platforms to get ideas from outside. If a language model can generate thousands of ideas in an hour, what is the crowd still for, and how do firms tell good ideas from fluent ones?
Papers in the last 30 days — open innovation: 2 · crowdsourcing: 1
Continuous-time models built on the Hamilton-Jacobi-Bellman equation are rising in the finance journals, while neural operators and physics-informed networks that solve such equations are growing in AI. All five recent papers that pair the two are in AI journals. Can these solvers make richer portfolio and asset-pricing models usable?
Papers in the last 30 days — Hamilton-Jacobi-Bellman equation: 1 · physics-informed neural networks: 8
Narcissism, CEO characteristics and CEO personality appear on 42 recent papers, mostly in the management and accounting journals, with measures built from signatures, photos and pronoun counts. Language models can now score personality from thousands of hours of speech. Do the old measures and the new ones agree, and does personality predict decisions out of sample?
Synthetic data generation appears on 77 recent papers and synthetic data on 18, almost all in AI journals, while model collapse, the degradation of models trained on their predecessors' output, appears on 3. The web is filling with generated text and images that will train the next models. How fast does quality decay, and what does that do to the value of human data?
Papers in the last 30 days — synthetic data: 1
Mixture-of-experts models are among the fastest-growing topics in the AI journals, and regime-switching models have a long history in finance and economics. A gate that picks experts works much like a regime switch, yet no finance or economics paper in the archive carries the keyword. Can it forecast better across market regimes?
Papers in the last 30 days — mixture of experts: 11 · regime switching: 4
CEO and managerial overconfidence appear on 98 recent papers, mostly measured from option-exercise behaviour or press coverage, as in the 2000s. Only 6 of them carry a text, language or speech keyword, and 4 of those are in AI journals. Can a language model score overconfidence from earnings calls, and does it predict the same decisions?
Market manipulation, stock market manipulation, price manipulation and pump-and-dump schemes appear on about 20 recent papers, split between the AI and finance journals. Spoofing, wash trading and coordinated pumps now happen in milliseconds and across venues, including crypto. Can detectors trained on known cases find new ones, and what do regulators miss?
Soft information and relationship lending appear on 26 recent papers in the finance and economics journals, but only 5 of them carry an algorithm, machine-learning, fintech or automation keyword. As banks automate small-business lending, what is lost when the loan officer's judgment goes, and can models learn it?
Papers in the last 30 days — soft information: 1
Price competition is rising steeply in the operations research journals and AI governance leads the AI journals, while collusion remains a small topic that is rarely tied to algorithms. When sellers let language-model agents set prices, do prices drift up without any agreement?
Papers in the last 30 days — collusion: 3 · price competition: 6
Confidence calibration and uncertainty calibration appear on 54 recent papers, mostly in AI journals, and 35 calibration papers involve language or vision-language models, with 94 calibration-related papers in the last quarter against about 57 per quarter before. When a model answers a financial or legal question, its stated confidence rarely matches its accuracy. Can calibration be fixed, and do users read it?
Papers in the last 30 days — uncertainty calibration: 1
Evaluation metrics appear on 56 recent papers and LLM-as-a-judge on 16, all in AI journals, with 25 evaluation papers in the last quarter in line with earlier quarters. Models now grade other models' answers, and firms use them to score customer conversations and reports. How reliable and how biased are the model judges, and how far do the biases propagate?
Papers in the last 30 days — evaluation metrics: 4 · LLM evaluation: 2
Agent-based modelling is rising in the operations research journals, and firms are starting to let AI agents place orders. The bullwhip effect, where small swings in demand grow larger up the supply chain, appears on only 9 recent papers, and just one of them carries an agent or AI keyword. Do AI ordering agents dampen the bullwhip, or amplify it when they all react to the same signals?
Papers in the last 30 days — bullwhip effect: 2 · agent-based modeling: 2
Benders decomposition, column generation, branch-and-cut and related methods appear on 240 recent papers, almost all in the operations research journals, and the topic is fading, with 42 papers in the last quarter against about 64 per quarter before. Only 16 of them use learning. Can learned cut selection, column pricing and branching make these methods faster on real problems?
Papers in the last 30 days — Benders decomposition: 4 · column generation: 3
Survey experiments are a staple of economics, and some researchers now simulate respondents with language models ("silicon samples"). Only 6 of 179 recent survey papers carry a language-model keyword, 5 of them in AI journals. Where do simulated answers match real people, and where would they mislead economic research?
Papers in the last 30 days — survey experiment: 1 · large language models: 95
Matheuristics and local search are among the fastest-growing keywords in the operations research journals, and recent systems let language models write and evolve heuristic code. Only 2 of the 252 heuristics papers in OR journals carry a language-model keyword. Can generated heuristics beat hand-designed ones on standard routing and scheduling instances?
Papers in the last 30 days — matheuristic: 3 · local search: 3
Knowledge transfer, sharing and management appear on 259 recent papers across the management and AI journals, but only 48 of them carry an AI or digital keyword, and knowledge hiding on 3. Firms are building assistants trained on employees' documents and chats. Do employees share less when sharing feeds a machine?
Papers in the last 30 days — knowledge sharing: 1
Knowledge graphs appear on 141 recent papers and neuro-symbolic AI on 14, almost all in AI journals; only 34 of 481 such papers concern accounting, tax, legal, regulatory, financial or audit uses. Compliance is a domain of explicit rules and entities, which language models alone handle badly. Do graphs and symbolic rules make them reliable enough for audits and tax?
Papers in the last 30 days — knowledge graphs: 9 · neuro-symbolic AI: 2
The technology acceptance model is one of the fastest-growing keywords across the tracked journals, and a quarter of recent TAM and UTAUT papers now study AI. But the model was built for workplace software decades ago. Do perceived usefulness and ease of use predict who actually uses generative AI, and what do they miss?
Papers in the last 30 days — technology acceptance model: 4 · UTAUT: 1
Aerial imagery is rising steeply in the AI journals, and economists already use night-time lights to measure growth. Yet none of the papers on the informal economy uses either. Can image models measure informal activity that surveys miss?
Papers in the last 30 days — informal economy: 2 · nighttime lights: 1
Human-AI collaboration is rising in the management and AI journals, while audit quality remains a core accounting topic. Few papers connect them. When auditors work with AI tools, does audit quality improve, and who answers for the judgments the tools shape?
Papers in the last 30 days — audit quality: 3 · human-AI collaboration: 8
Financial contagion and systemic risk are rising in the finance and economics journals, and graph neural networks are among the largest topics in the AI journals. The two have met in only five papers. Can a graph model learn the channels of contagion that network models fix in advance?
Papers in the last 30 days — financial contagion: 12 · systemic risk: 12
Accounting education is one of the fastest-growing keywords in the accounting journals, and a quarter of those papers now involve generative AI. Does an AI tutor help students learn accounting, or only finish the homework?
Papers in the last 30 days — accounting education: 2 · generative AI: 15
Employment is rising as a keyword in the economics journals, and generative AI is growing there fast. Yet of nearly 1,200 recent papers on employment, wages and labour markets, only 14 carry a generative-AI keyword. Is AI taking the tasks that junior workers used to learn on, and what happens to their careers?
Papers in the last 30 days — young workers: 1 · employment effects: 3
Foundation models are rising fast in the AI journals, and time-series versions now forecast series they were never trained on. Economists forecast inflation and growth with surveys and small models. Do the new models beat the Survey of Professional Forecasters in real time?
Papers in the last 30 days — foundation models: 6 · inflation forecasting: 1
Contrastive learning is one of the largest topics in the AI journals, yet only 6 of 749 recent contrastive-learning papers carry a finance or firm keyword. Finance research still defines peer firms mostly by industry codes. Would firm embeddings learned contrastively from filings, returns and supply chains give better peers for valuation, benchmarking and pay?
Papers in the last 30 days — contrastive learning: 27 · peer firms: 1
Investor sentiment and sentiment analysis appear on 490 recent papers across the AI and finance journals, but only 22 of them carry a multilingual, cross-lingual or non-English language keyword, and financial sentiment analysis on 6. Most of the world's markets are discussed in languages the standard tools were not built for. How well do the tools work there, and do they move markets?
Papers in the last 30 days — financial sentiment analysis: 1 · sentiment analysis: 6
Misinformation, fake news detection and rumor detection appear on 185 recent papers, 140 of them in AI journals, and only 10 carry a market, stock, investor or finance keyword. Fake earnings leaks, forged press releases and coordinated rumours now move prices within minutes. How much do they move, who profits, and can detection work at trading speed?
Papers in the last 30 days — misinformation: 2 · fake news detection: 6
Federated learning, which trains one model across organizations without pooling their data, is a large topic in the AI journals, yet only 7 of 398 recent papers on it carry a finance, banking, fraud or money-laundering keyword. Money launderers spread transactions across banks precisely because no single bank sees the whole pattern. Can banks catch them together without sharing customer data?
Papers in the last 30 days — federated learning: 23 · anti-money laundering: 1
Online reviews appear on 54 recent papers, and 17 of 73 review papers carry a fake, manipulation or detection keyword. Fake reviews were once cheap but detectable; language models make them cheap and fluent. Can platforms still tell, how much of the review stock is now synthetic, and do consumers still trust ratings?
Papers in the last 30 days — online reviews: 7
Fairness and algorithmic bias are growing in the AI journals, and credit scoring is one of the areas where AI rules are strictest. Yet only 19 of 473 recent fairness and bias papers carry a credit or lending keyword. What would an audit that lenders and regulators can both use look like, and what does fairness cost in default risk?
Papers in the last 30 days — algorithmic fairness: 2 · credit scoring: 2
Class imbalance is one of the largest topics in the AI journals, and fraud detection is where it matters most for finance. But only 10 of 174 recent fraud papers carry a cost, profit or loss keyword; most report accuracy or F1 scores. Would ranking fraud models by the money they save change which ones win?
Papers in the last 30 days — fraud detection: 3 · class imbalance: 22
Industrial and infrastructure inspection appear on 139 recent papers, almost all in AI journals about detecting defects in images, and 25 of them involve drones or robots. Ageing bridges, pipelines and power lines need inspecting more often than inspectors can manage. How much do drone inspections cost and miss compared with people, and do they change maintenance decisions?
Papers in the last 30 days — industrial inspection: 3
Explainable AI is one of the largest topics in the AI journals and is spreading into management, finance and economics. Explanation fidelity (whether an explanation reflects what the model really does) is a fast-rising but still small keyword. When SHAP, LIME and attention maps disagree about the same prediction, which should a loan officer or a doctor believe?
Papers in the last 30 days — explainable AI: 29 · explanation fidelity: 1
Information transparency and corporate disclosure are rising in the accounting journals, and language models now read filings for investors. Yet few disclosure papers mention language models or XBRL. Is disclosure written for machines changing what transparency means?
Papers in the last 30 days — information transparency: 7 · corporate disclosure: 14
Earnings management is a mainstay of the finance and accounting journals, still measured mostly with accrual regressions such as the modified Jones model. Only 4 of 156 recent papers on it carry a machine-learning or text keyword. Do learned models, including language models reading the filings, flag manipulation that accrual models miss?
Papers in the last 30 days — earnings management: 6 · accrual-based earnings management: 1
Demand and sales forecasting appear on 66 recent papers in the operations research and AI journals, but only 9 of them carry a language-model, foundation-model or text keyword. Retailers now feed product descriptions, reviews, weather and news into forecasting systems. Does text improve forecasts for new products, where history is missing, and does it help where history is long?
Papers in the last 30 days — demand forecasting: 10 · sales forecasting: 1
Unethical behavior and moral disengagement are rising in the management journals, and AI tools now act on people's behalf. Recent experiments suggest people cheat more when a machine does the act for them. Does that happen at work, with sales targets, expense claims or performance reports?
Papers in the last 30 days — unethical behavior: 1 · moral disengagement: 2
Deepfake detection appears on 45 recent papers and AI-generated image detection on 13, almost all in AI journals, with 23 papers in the last quarter against about 21 per quarter before. Detectors score well on the datasets they were built with and fail on the next generator. How well do they work on real fraud, elections and markets, and who should run them?
Papers in the last 30 days — deepfake detection: 1
Reinforcement learning appears on more than 700 recent papers, and 41 of them apply it to portfolios, trading or hedging, almost all in AI and operations research journals. The results are rarely reproduced and rarely compared with simple rules after costs. Which of the reported gains survive replication on new data?
Papers in the last 30 days — deep reinforcement learning: 17 · portfolio optimization: 13
Data leakage appears on 16 recent papers and memorization on 3, almost all in AI journals, while thousands of papers report results on public benchmarks that language models may have seen in training. How much of reported progress is contamination, and can finance and economics benchmarks be protected?
Papers in the last 30 days — data leakage: 2
Data centres appear on just 16 recent papers across the tracked journals, though the count is rising, and electricity demand on 17. AI training and inference are driving the first sustained growth in electricity demand in rich countries for decades. Who pays for the new capacity, and do data centres crowd out electrification elsewhere?
Lobbying, corporate political activity and campaign contributions appear on 49 recent papers, mostly about tax, trade and environmental policy, while AI regulation appears on 53 papers as a question of design. The EU AI Act and dozens of US state bills attracted record lobbying by technology firms. What did the money change in the texts?
Papers in the last 30 days — corporate political activity: 1
Copyright protection, data ownership, content creators and the creative industries appear on 96 recent papers across the AI, management and economics journals, and 37 of them involve generative AI. Courts are deciding whether training on copyrighted work is fair use, and licensing deals are being signed. What are creators losing or gaining, and what would a licensing market look like?
Papers in the last 30 days — copyright protection: 1 · data ownership: 1
Coopetition, cooperation between competitors, appears on 31 recent papers split between the management and operations research journals, with 7 in the last quarter, in line with earlier quarters. AI firms compete fiercely while sharing open models, benchmarks and safety research, and renting computing from each other. When does sharing help a rival more than oneself?
Papers in the last 30 days — coopetition: 2
Distribution shift, concept drift and non-stationary time series appear on 371 recent papers, almost all in AI journals, and only 27 of them concern credit, fraud, banking or markets. Credit and fraud models decay as borrowers, fraudsters and the economy change. How fast, how should it be detected, and when should a model be retrained rather than recalibrated?
Papers in the last 30 days — concept drift: 6 · distribution shift: 8
Code generation, program synthesis, vulnerability detection and software development appear on 75 recent papers, mostly in AI journals, and 35 of them touch productivity, quality or security. Coding assistants are the most widely adopted generative AI tool at work. What happened to developer output, code quality and the junior developer's job?
Customer satisfaction and customer engagement are among the fastest-growing keywords in the management journals, and firms keep replacing service agents with chatbots. The EU AI Act requires telling people when they are talking to an AI. Does disclosure lower satisfaction, and can good service recovery make up for it?
Papers in the last 30 days — chatbots: 6 · customer satisfaction: 7
Double machine learning, causal forests, conditional average treatment effects and policy learning appear on 50 recent papers, split between the AI and economics journals, with 14 in the last quarter against about 11 per quarter before. The methods promise to find who benefits most from a programme. Do the targeting rules they produce hold up in new samples, and are they fair?
Papers in the last 30 days — double machine learning: 2
Edge and cloud computing appear on 151 recent papers, with 43 in the last quarter against about 34 per quarter before, and 46 of them carry an energy, carbon or green keyword, mostly about device energy. Carbon-aware scheduling, which moves computing jobs to times and places where the grid is clean, appears on one paper. How much carbon could it save, and who would pay for the delay?
Papers in the last 30 days — cloud computing: 2 · edge computing: 4
Customer service and call centres appear on 25 recent papers, and 10 of them carry a chatbot, automation or language-model keyword. Firms now route calls through voice agents that handle part of each conversation and hand the rest to people. How should queues, staffing and handovers be designed when the first responder is a model?
Papers in the last 30 days — customer service: 3
Distributionally robust optimization is one of the fastest-growing methods in operations research, and uncertainty estimation is rising in the AI journals, yet only two papers in the archive touch both. Can calibrated uncertainty from machine learning decide how much robustness an optimization model buys?
Papers in the last 30 days — distributionally robust optimization: 9 · conformal prediction: 2
The dataset ranking shows CIFAR and MNIST still among the most-mentioned image benchmarks in AI journal papers, while cross-dataset generalization and reproducibility are rising keywords. Do gains on these old benchmarks carry over to harder, shifted data?
Papers in the last 30 days — benchmark dataset: 6 · cross-dataset generalization: 2
Autonomous driving and autonomous vehicles appear on 314 recent papers, 270 of them in AI journals about perception and control, and only 29 carry a liability, insurance, regulation, acceptance or ethics keyword, with liability on 2. Driverless cars are now on public roads. When one crashes, who pays, and does the answer shape what gets deployed?
Papers in the last 30 days — autonomous driving: 5 · autonomous driving safety: 1
Annotation cost reduction and reinforcement learning from human feedback appear on 44 recent papers in the AI journals, among 71 papers on annotation and human feedback, but only 8 of them carry a worker, labour, wage, platform or ethics keyword. Millions of people label data and rate model outputs, often through platforms in low-income countries. What are they paid, and how does the pay shape the data?
Analyst attention is rising in the finance journals, and AI tools now summarize filings and earnings calls for anyone. Only 13 of 246 recent papers on analysts carry an AI keyword. Do analysts drop the small firms that machines now summarize, or use the time saved to dig deeper?
Papers in the last 30 days — analyst attention: 1
Recruitment, job matching and hiring discrimination appear on 109 recent papers across the economics, management and AI journals, with 32 in the last quarter against about 25 per quarter before; 32 of them carry a bias, fairness, algorithm or AI keyword. Screening tools now read millions of applications, and New York requires bias audits of them. What do the audits find, and do they change outcomes?
Trust in AI appears on 76 recent papers, algorithm aversion on 10 and algorithm appreciation on 3, across the AI and management journals, with 28 papers in the last quarter against about 25 per quarter before. People once rejected algorithmic advice after seeing it err. Language models that explain themselves fluently may get too much trust instead. Which way does it go, and for whom?
Papers in the last 30 days — algorithm aversion: 1 · trust in AI: 5
Charitable giving and crowdfunding appear on 168 recent papers, with 44 in the last quarter against about 38 per quarter before, but only 23 of them carry an AI, text, narrative or image keyword. Campaign success depends on the story. If fundraisers now write with language models, do donors give more, and do they give to the right causes?
Papers in the last 30 days — charitable giving: 4 · crowdfunding: 3
Weather forecasting, precipitation forecasting and typhoons appear on 45 recent papers, mostly in AI journals, and only 9 carry an economic, energy, agricultural, insurance or market keyword. Machine learning models now beat the physics-based systems on many measures at a fraction of the cost. Who captures the value, and does it reach the farmers, utilities and insurers who need it?
Papers in the last 30 days — weather forecasting: 1
Artificial intelligence and stock price crash risk are both rising in the finance journals, and "AI washing" has appeared as a keyword. When the AI story in disclosures runs ahead of actual AI investment, does the stock carry more crash risk?
Papers in the last 30 days — AI washing: 1 · stock price crash risk: 6
Purchase intention, purchase decisions and consumer choice appear on 178 recent papers, mostly in management and economics, but only 14 of them carry a generative-AI, conversational-AI or recommendation keyword. Shoppers are starting to let AI assistants search, compare and buy. Who gets recommended, and what happens to brands and prices when a model makes the choice?
Papers in the last 30 days — purchase intention: 13 · online shopping: 2
Patents appear on 377 recent papers, and patent quality on 24, but only 32 of the patent papers carry an AI or software keyword, and "AI patents" on 14. AI patent filings have exploded, especially in China, while much AI progress is published as open code. What do AI patents protect, and do they predict anything?
Papers in the last 30 days — AI patents: 1 · patent quality: 1
AI regulation, trustworthy AI, responsible AI and the EU AI Act appear on 229 recent papers, mostly in AI and management journals, but only 12 of them carry an insurance, liability, compensation or damages keyword, and AI liability on 2. Rules now assign duties; they say little about who pays for harm. Can AI harms be insured, and would insurance discipline deployment?
Papers in the last 30 days — AI regulation: 4 · EU AI Act: 1
Peer review, research productivity and academic writing appear on 56 recent papers across the economics and AI journals, with 17 in the last quarter against about 12 per quarter before, and 17 of them involve language models. Journals report AI-written reviews and AI-written submissions. How much of each is there, and does it change what gets published?
Papers in the last 30 days — peer review: 3 · academic writing: 1
The keyword "AI governance" is growing fastest across the tracked journals, while corporate governance keeps rising in economics and management. Is board-level oversight of AI turning into a research stream of its own?
Papers in the last 30 days — AI governance: 9 · corporate governance: 51
Advertising appears on 15 recent papers, online advertising on 9 and AI-generated advertising on 3, in the management and economics journals, while AI-generated content appears on 33. Brands now generate images, copy and whole campaigns with models. Does disclosure of AI use change how consumers respond, and does generated advertising work?
Papers in the last 30 days — advertising: 2 · AI-generated content: 1
Outsourcing, offshoring and digital trade appear on 104 recent papers, but only one of them carries an AI, remote-work or online-labour keyword. If language models make call centres, coding and back-office work cheaper to automate than to offshore, the countries that built service exports on cheap labour face a new competitor. Is service offshoring already slowing?
Papers in the last 30 days — digital trade: 2 · outsourcing: 1
Firms spend heavily on data, software and AI models, but much of that spending is expensed and barely shows on balance sheets. Intangible assets are a steady finance topic, yet only one of 57 recent papers on them carries an AI, data or software keyword. How large is the missing AI capital, and does it distort book-based measures?
Papers in the last 30 days — intangible assets: 3 · intangible capital: 2
Transaction costs are back as a growing keyword in the management journals, while agentic AI and language-model agents rise in AI. Coase argued that firms exist because using markets is costly. If AI agents can search, negotiate and monitor contracts cheaply, do firms buy more and make less?
Papers in the last 30 days — transaction cost economics: 3 · agentic AI: 3
Absorptive capacity tops the list of growing keywords across the tracked journals, while AI adoption keeps rising. Do firms need prior know-how to profit from AI, so that AI widens the gap between leaders and laggards?
Papers in the last 30 days — absorptive capacity: 9 · artificial intelligence: 37
In the finance journals both "artificial intelligence" and "price discovery" gained papers this year, but rarely in the same paper. Do markets with more AI-assisted trading take in news faster, or just louder?
Papers in the last 30 days — artificial intelligence: 37 · price discovery: 7
Scholar4AI. (2026). Research ideas tagged “ai”. Journal Research Rankings. Retrieved October 26, 2026, from https://scholar4ai.com/ideas/tags/ai
@misc{scholar4ai-ideas-tags-ai,
author = {{Scholar4AI}},
title = {Research ideas tagged “ai”},
year = {2026},
howpublished = {\url{https://scholar4ai.com/ideas/tags/ai}},
urldate = {2026-10-26},
note = {Journal Research Rankings; data from OpenAlex (CC0); licensed CC BY-NC 4.0}
}