Understand the signal before you apply.
AI releases, research, funding shifts and application guidance—checked against cited sources and connected to current global opportunities.
213 published insights · page 4 of 12
OpenAI updates Model Spec on teen interactions, unsupported premises and capability transparency
OpenAI's August 18, 2026 Model Spec update adds clearer guidance for teen relational interactions, false or unsupported premises, and communicating assistant capabilities and limits.
OpenAI backs 14 policy projects on economic opportunity and resilience in the Intelligence Age
OpenAI is funding 14 independent projects that will test policy ideas for broadening economic opportunity and strengthening societal resilience as advanced AI capabilities spread.
OpenAI and CodeAI Launch Student AI-Literacy Partnership Focused on Critical Use
OpenAI and CodeAI are partnering on AI literacy, youth safety, classroom resources and student builder programs designed to teach young people how to question and responsibly use AI.
NVIDIA AVO Shows How Agent Architecture Can Unlock Long-Horizon AI Performance
NVIDIA reports that its Agentic Variation Operators system completed the full public ARC-AGI-3 set while also demonstrating autonomous GPU-kernel optimization, highlighting the importance of...
AWS Vector Solutions: Building Agentic AI Where Enterprise Data Already Lives
AWS argues that many agentic-AI retrieval architectures can reduce complexity by adding vector capabilities to the services where enterprise data already resides.
AWS makes AgentCore payments generally available for transacting AI agents
Amazon Bedrock AgentCore payments is now generally available, giving AI agents a managed way to discover paid services, authorize transactions and enforce spending controls without exposing...
OpenAI Expands Zero Data Retention for Frontier Models and Previews Private Safety Processing
OpenAI says eligible API customers can use frontier models with Zero Data Retention while it tests a privacy-preserving approach for detecting safety risks across related interactions.
OpenAI launches AI Futures to examine rights, agency and institutions in the AI era
OpenAI has launched AI Futures, a Strategic Futures initiative focused on how societies can preserve individual rights, agency and accountable institutions as advanced AI changes economic an...
OpenAI launches $5M initiative for oversight of AI in national security
OpenAI has announced a new initiative to help authorized democratic oversight bodies build expertise and tools for reviewing government use of AI in national-security settings.
Anthropic Fellows Research Shows Autonomous AI Agents Can Run Weak-to-Strong Alignment Experiments
Claude-powered automated alignment researchers proposed ideas, ran experiments and iterated on weak-to-strong supervision, reaching strong benchmark gains while also exposing reward-hacking...
Google DeepMind brings multilingual sign-language-to-text AI to Pixel 11
Google DeepMind's new SL2T model moves sign-language translation into consumer products, starting with ASL-to-English dictation in Gboard and Live Transcribe on Pixel 11.
Microsoft MAI-Code-1.1-Flash brings faster, cheaper coding AI to GitHub Copilot
Microsoft says its MAI-Code-1.1-Flash coding model improves token efficiency and cost while moving into production use in GitHub Copilot.
Google DeepMind introduces Gemini 3.7 Flash for coding and agent workflows
Gemini 3.7 Flash adds stronger reasoning for coding and agents, a 1M-token context window, 64K output and configurable thinking controls.
NVIDIA Nsight AI combines CUDA MCP access, self-hosted Copilot and performance guidance
NVIDIA Nsight AI gives coding agents current CUDA knowledge through a hosted MCP server, adds CUDA-aware Copilot workflows, and offers a self-hosted blueprint for GPU development teams.
OpenAI slows frontier model scaling as cyber-critical capability risks rise
OpenAI says it temporarily slowed frontier scaling and tightened monitoring, alignment and research-environment security after cyber capability concerns.
Anthropic Research Finds Fine-Tuned AI Lie Detectors Struggle to Generalize
New Anthropic-linked alignment research reports that lie detectors trained on specific deception patterns performed well in-distribution but transferred poorly to unseen lie types.
OpenAI's Defender's Window: Why Security Teams Need AI Agents Now
OpenAI argues that rapidly improving cyber-capable AI is shrinking defenders' response window and recommends faster adoption of agentic security workflows alongside stronger security fundame...
Anthropic's CVD Dashboard Shows the Scale—and Bottlenecks—of AI-Assisted Vulnerability Discovery
Anthropic's August 26 coordinated-vulnerability-disclosure snapshot reports 2,300 disclosed findings across 392 open-source projects, highlighting both the reach of AI-assisted security rese...
Turn intelligence into applications.
Create a free alert for the topics, roles or countries that matter. We email only new verified matches.
More ways to save
Discover deals, coupons and free courses on our sister site.
