The agency has created a dedicated centre for embodied-AI development, while its first real operational humanoid remains a 2028 target rather than a deployed capability.
AI News Bank
A continuously updated, source-transparent newsroom for AI news, deals, companies, learning, product changes, rumors, and expectations.
589 posts399 source domains
Everything AI, separated by evidence type.
203
News
Confirmed reporting and analysis across models, policy, safety, infrastructure, and society.
25Deals
Funding, M&A, strategic investments, capacity commitments, and market structure.
27Companies
Platform moves, pricing, partnerships, competitive strategy, and business model shifts.
51Learn
Research artifacts, explainers, resources, benchmarks, and practical learning coverage.
242Changes
Model releases, access changes, APIs, deprecations, product updates, and availability.
7Rumors
Clearly labeled reports and unconfirmed signals with source-quality caveats.
34Expectations
Forecasts, deadlines, upcoming releases, and next checkpoints — never framed as fact.
Newest sourced posts
The research framework converts failed agent traces into evidence-backed constraints, but its public records disagree on benchmark size.
Workspaces are expected to keep their existing enabled or disabled value when the control disappears, making the period before removal the clearest time for administrators to verify and document the setting.
Developers who used the managed agent-loop endpoint are being directed to Databricks Apps, where code, server behavior, deployment and identity choices become explicit parts of the application.
The native SQL operation ranks matches for every query row in exact or approximate mode, bringing offline vector workloads into the same engine that holds the source tables.
Organization and enterprise security teams can now audit effective AI Scan enablement by repository, filter the coverage view and export the result, although a not-enabled state does not explain its cause.
The four-principle paper asks organizations to match control, choice, flexibility and resilience to each workload, while offering no independent outcome evidence.
The service pairs infrastructure and IaC analysis with prioritized remediation packages, but AWS has not published independent performance evidence.
The response opens a new AI Airlock phase and sets lifecycle monitoring, draft guidance and a Spring 2027 roadmap as the next implementation checkpoints.
The new experience turns recipes, photos, lists and meal requests into a Safeway cart before handing shoppers to the retailer's checkout.
The company says the certification covers ElevenAgents and speech APIs only under zero-retention and U.S.-residency settings, narrowing the path into non-sensitive federal workloads.
The prototype combines solar-wind forecasts with local geology and grid data to estimate substation-level exposure, but it still needs utility validation before operational use.