The AI Street Journal for 8 September 2026
The two publishers have joined the growing line of news organisations taking copyright claims over AI training into court. The case adds pressure even as Microsoft says it wants a discussion.
The briefing
A fresh copyright suit adds to OpenAI and Microsoft’s legal pile-up, while two arXiv papers sketch more capable finance forecasting and a sterner framework for judging AI in hiring. One is a court fight, two are preprints, and all three come with caveats attached, as they should.
Seattle Times and Newsday sue OpenAI and Microsoft
Seattle Times and Newsday filed a lawsuit against OpenAI and Microsoft over alleged use of their journalism for AI training, according to TechCrunch. Microsoft said it was surprised by the suit.
The Seattle Times and Newsday have filed a lawsuit against OpenAI and Microsoft over the alleged use of their journalism to train AI systems, according to TechCrunch. The complaint places two more publishers alongside earlier claimants, including The New York Times, in a widening copyright dispute over training data and AI-generated outputs.
TechCrunch reports that the lawsuit argues generative AI could damage the news business by relying on publishers’ work while competing with it. The article also notes an awkward detail: Microsoft and OpenAI have funded some Seattle Times journalism projects and fellowships, which gives the dispute an extra layer of small-room tension.
What changes
In practical terms, the case increases the legal and commercial pressure on OpenAI and Microsoft as they continue to defend how AI systems are trained and what they reproduce. It also signals that the litigation is no longer confined to a single marquee newspaper or a narrow group of rights holders.
The limits are plain enough. TechCrunch summarises the filing, not the defendants’ full legal response, and the article does not show the complaint itself. Microsoft, via a spokesperson quoted by GeekWire and relayed by TechCrunch, said it was surprised by the lawsuit and open to exploring solutions. The suit’s merits, any evidence testing, and any settlement or ruling remain to be seen.
More publisher suits mean higher legal risk for model providers and their partners. For businesses using AI search, assistants or content tools, the practical issue is not courtroom theatre but whether licensing costs, product design or output restrictions change as these cases accumulate.
Market signal
Authors describe EXAONE Finance forecasting model
An arXiv preprint introduces EXAONE Finance, a financial forecasting model using linear-time operators instead of self-attention. The authors claim leading benchmark results, but the report is not peer reviewed.
A technical report on arXiv presents EXAONE Forecast for Finance, described by its authors as a financial time-series foundation model built for forecasting across equities, foreign exchange, commodities, crypto-assets, fixed income and macroeconomic indicators. The paper says existing time-series foundation models are often aimed at general data and can struggle with the long, many-variable and intermittently missing sequences common in finance.
To address that, the authors replace self-attention with two linear-time components: a causal 1D convolution for temporal mixing and a group-aware pooling multilayer perceptron for mixing across variables. They also describe a masked context augmentation method intended to make the model more robust when chunks of market data are missing.
Claims and limits
The authors say the model was pretrained on a large-scale financial corpus and ranks first on all three tiers of the FinVerse benchmark: point-forecast accuracy, cross-sectional asset ranking and portfolio profitability. If those results hold up, the design could matter for firms that want broader market coverage without the heavier compute costs associated with some attention-based systems.
The limits are important. This is an arXiv preprint and technical report, not a peer-reviewed paper. The evidence here is the abstract alone, so details such as training data construction, benchmark design, implementation choices and comparison baselines are not independently tested in the supplied material. In short, promising on paper, which is where many things are impeccable.
For trading desks, researchers and vendors, the practical attraction is efficiency under awkward real-world data conditions, not just another benchmark line. But anyone treating the claimed ranking or profitability results as settled fact would be getting ahead of both peer review and replication.
What to watch
Recruitment AI review urges broader evaluation rules
A narrative review on arXiv says AI recruitment systems now span multi-stage workflows and agents, while evidence on fairness, privacy and security remains incomplete within its coded set.
A systematized narrative review on arXiv argues that AI in recruitment has shifted from matching candidates and vacancies toward multi-stage workflows that retrieve evidence, compare applicants and can support or carry out actions. The authors say they organised 40 representative works, alongside industrial and legal sources, using a search and coding process updated through late July 2026 with targeted updates into early September.
Their central claim is that evaluation methods have lagged behind the systems themselves. The review traces a move from similarity scoring to reciprocal suitability, from a single model to a compound workflow, and from offline prediction toward evaluation tied to evidence and productivity.
What the review found
The authors say persistent gaps arise because behavioural labels can mix up exposure, preference and qualification; private or synthetic datasets can limit external validity; and final output scores can hide failures inside a pipeline. Within their coded set, they report that privacy was not directly evaluated and that no row jointly evaluated utility, fairness, privacy and security.
Those findings come with explicit limits. The paper says it is not a prevalence estimate of the whole field and that its observations describe the coded set rather than every hiring product or study. It is also a preprint, not peer-reviewed evidence. Even so, for employers and vendors using AI in recruitment, the paper offers a more disciplined way to ask what a system actually proves, as opposed to what a dashboard implies.
Hiring tools affect access to jobs, appeals and compliance risk. A framework that separates modest evidence from grander claims could help employers ask better procurement questions, and help regulators or auditors spot where performance claims outrun what has actually been tested.
What to watch next
- More publisher suits mean higher legal risk for model providers and their partners. For businesses using AI search, assistants or content tools, the practical issue is not courtroom theatre but whether licensing costs, product design or output restrictions change as these cases accumulate.
- For trading desks, researchers and vendors, the practical attraction is efficiency under awkward real-world data conditions, not just another benchmark line. But anyone treating the claimed ranking or profitability results as settled fact would be getting ahead of both peer review and replication.
- Hiring tools affect access to jobs, appeals and compliance risk. A framework that separates modest evidence from grander claims could help employers ask better procurement questions, and help regulators or auditors spot where performance claims outrun what has actually been tested.
The takeaway
More publisher suits mean higher legal risk for model providers and their partners. For businesses using AI search, assistants or content tools, the practical issue is not courtroom theatre but whether licensing costs, product design or output restrictions change as these cases accumulate.
The editor’s view
For trading desks, researchers and vendors, the practical attraction is efficiency under awkward real-world data conditions, not just another benchmark line. But anyone treating the claimed ranking or profitability results as settled fact would be getting ahead of both peer review and replication.