Coding agents now write at a scale humans cannot review line by line. This week brings a formal verification tool from Microsoft Research, a benchmark showing long-horizon failure, and evidence that agent-written tests can run while being wrong.
Franklin Templeton is developing AI workflows to help portfolio managers compare views and examine risks, using OpenAI's Codex. Rogo Intelligence reduced agent steps by 42% and input tokens by 48% across 8,800 tasks. Boosted.ai's agents monitor earnings and build models before analysts ask.
SoundHound reported Q3 revenue of $62.4 million, up 48.5% year over year, and guides to sustained adjusted EBITDA profitability within the second half of 2027. Its LivePerson acquisition closed on September 4, 2026.
Four preprints this week move AI safety research past prompt jailbreaks: alignment can collapse under KV cache quantisation, chain-of-thought monitors can be fooled by covert misalignment, and self-evolving agents can drift unsafe on their own.
Four preprints this week attack the same bottleneck in embodied AI: robots need more experience than anyone can collect by hand. The answers are human video at scale, explicit memory, spatial awareness and self-improvement.
Waymo is preparing for winter with sensor heaters and snow tires. Zoox signed a multi-year Formula 1 partnership. May Mobility and Forterra advanced to Phase III of a Pentagon challenge worth up to $30 million.
NVIDIA says agentic throughput per dollar rose up to 30x between Blackwell Ultra and Vera Rubin on SemiAnalysis AgentX. Cerebras says TSMC qualifies its wafers for 10 years at 105°C. Broadcom says a 1.5x bandwidth-delay-product buffer keeps multi-site training near-ideal.
Kalepa tripled revenue and customer count in a year, according to the company. Five Sigma's customers report 60% cuts in routine claims work. Mortgage AI faces a harder market: rates above 7% and falling origination volume.
Four preprints this week audit the evaluation pipeline itself: how quality is judged, how agent restarts are decided, how benchmarks are compressed, and how work is routed between agents. The results are uncomfortable for anyone who trusts a leaderboard.
Talkdesk's survey puts the maturity gap in numbers: adoption is near-universal, cross-department execution is rare, and only 5% of companies can quantify the business impact.
Inductive placed first in the OpenADMET CYP Inhibition Challenge, its fourth consecutive first-place finish. Deep Genomics' DeepRNAi uses biological foundation models to predict siRNA activity. Chai Discovery entered a collaboration with Roche.
Hippocratic AI released the AI Compliance Supervisor, which monitors every call rather than a sample. Viz.ai's Viz HCM algorithm flagged patients with a 2.68-fold higher hazard of adverse clinical events. Saama says database lock delays cost $780,024 to $2.34 million.
APT, GPUStack, the MCP Inspector and OpenSSH all disclosed serious flaws within days. One was already exploited in the wild. The fixes are simple; the exposure is wide.
TinyBrowser charges $0.002 per browser minute for managed browser infrastructure. Arcade deploys MCP servers into enterprise environments with one command. Braintrust added tracing so developers can see how agents use their tools.
No single protein language model wins across 32 tasks, homology-controlled splits collapse simple probes, and the week's new tools are built to be reused rather than ranked.
A materials model matches its rivals on 35 to 60 percent less data, a polymer review maps the road to autonomous design, and a 700-crystal dataset gives organic electronics a training set.
October volume hit $123.6 billion. Kalshi followed with regulated perpetual futures on single stocks. The question is whether these are markets or media.
A transplant risk score holds up but miscalibrates, early valve replacement cuts hospital stays without moving survival, and stroke survivors die of different causes than the stroke that brought them in.
Wastewater tracking at the World Cup, a four-year TB resistance survey in Morocco, nine years of respiratory infections in Lyon, and a one-tube TB typing method that multiplexes 96 samples.
Baltimore is finalizing a trade for Jacoby Brissett, Chicago's new defensive tackle dominated his debut, and Jonathan Taylor's future resolves by Tuesday at 4 p.m.
Two conditional loans back domestic production of chip wafers and cathode material, while troops and agents train against the drones those factories will have to face.
Connecticut's offshore wind farm switched on, California's new reservoir cleared its last permit, and Pennsylvania and New York moved money and megawatts.
Machine learning reads battery structures in seconds and models interfaces that used to take weeks, yet hydrogen storage in palladium hits a hard ceiling and the best interface models still get the energies wrong.