Realtime AI News
Harvey turns legal context into stronger drafts with GPT-6 Astra
OpenAI published a case story describing how the legal AI company Harvey uses GPT-6 Astra to produce more structured, context-aware legal documents. The pitch is that better drafting output frees lawyers to focus on strategy instead of document mechanics.
OpenAI published a case story on September 23 titled "Harvey turns legal context into stronger drafts with GPT-6 Astra," describing how the legal AI company Harvey applies GPT-6 Astra inside its drafting workflow. The logic is straightforward: feed the legal context of a matter into the model and get a stronger first draft back.
According to OpenAI, Harvey hands the legal context of a matter to GPT-6 Astra and lets the model produce the document from there, so the output reflects the specifics of the case at hand. OpenAI stresses that the resulting text is both more complete in structure and closer to the situation it is written for.
OpenAI frames the result with two descriptors: GPT-6 Astra produces more structured, context-aware legal documents. In legal drafting those are the hard parts, because facts, clauses and argument have to hold together inside one coherent document.
OpenAI frames the payoff in human terms: freed from drafting mechanics, lawyers can focus on strategy. That is the same promise the industry has been making for legal AI, now attached to a named model rather than a general claim about language models.
Why it matters: legal work is among the highest-value enterprise domains for large language models, and context awareness plus structured output are exactly the capabilities that decide whether a model stays a demo or becomes a production tool. By publishing a partner case rather than a benchmark chart, OpenAI points its capability marketing at real workflows.
What to watch next is where this capability stops being enough. Once drafting quality is no longer the bottleneck, competition in legal AI shifts toward verifiability, citation provenance and compliance review, the parts of the job that cannot be delegated to fluent text alone.
Why it matters
The case study anchors the value of LLMs in legal work on context awareness and structured output rather than raw generation speed. If that holds up, legal drafting moves from assisted writing toward production-grade workflows.
Nearby Updates
All09/23, 20:00
Sam Altman takes the AI safety case to the UN Security Council
OpenAI has published CEO Sam Altman’s remarks to the United Nations Security Council, covering AI safety, human control over AI systems and international cooperation. The appearance puts frontier-model risk on the agenda of the body responsible for international peace and security, a sign that AI governance is now treated as a cross-border policy question.
09/23, 20:00
OpenAI case study: invideo improves color grading 3x with GPT-6 Astra
OpenAI has published a customer story detailing how the video creation platform invideo uses GPT-6 Astra to plan edits with greater precision, reporting a threefold improvement in color correction and grading. The case study also credits the model with helping invideo produce 50 custom effects in a single day, a signal that frontier models are moving deeper into professional post-production work.
09/23, 20:00
Ema raises $77M as AI starts eating into enterprise software and services
Ema has raised $77 million in a new round, bringing its total funding to $140 million, and now counts more than 50 enterprise customers including Google and Microsoft. TechCrunch frames the round as evidence that AI agents are beginning to absorb spending that once went to enterprise software and services.
09/23, 19:54
OpenAI seeks independent reviews of its AI safeguards and serious incidents
OpenAI is seeking independent reviews of its AI safeguards and of serious incidents, according to a report by citybiz, addressing the question of who verifies frontier-model safety claims. The move would shift safety assessment from an internal process toward a more externally supervised framework.