AIImpactLab
9 impact clustersWork & marketsCognition & learningTechnical failuresPrivacySecurityEnvironmentLaw & informationSocial good & healthSystemic risk
Search the evidence

Find the signal.

Search titles, impact clusters, countries, organizations and the full text of every analysis.

5 stories found

A corporate AI token meter is compared with an employee profile, pull requests, performance scores, and a rapidly changing cost dashboard.
Work & marketsUnited States+4 clusters01

Rippling cut AI token costs by routing work. Now it wants to score employee ROI

Rippling says unchecked AI spending grew 80 percent month over month and put it on a path to spend 40 percent of its research-and-development headcount budget on tokens. The company found that roughly 10 to 15 percent of employees drove about 60 percent of total AI spend, with one engineer spending $50,000 in a month. It then capped tools, routed tasks through cheaper models, connected usage to work outputs, and says the projected burden fell to 10 to 15 percent of the headcount budget without reducing overall token use. Those are vendor-reported results, not independent evidence. The new AI Spend Console extends that logic to customers by mapping individual and team costs against pull requests, performance ratings, rework, and other outputs. Cost control is sensible. Turning token consumption and imperfect productivity proxies into employee scores requires strict purpose limits, transparency, and appeal.

5 min
Charcoal-and-lime conceptual illustration: a calculator rests above a large stack of correction papers.
Work & marketsGlobal+1 clusters02

The AI bill arrived. It included the work nobody counted.

The first AI bill is easy to see. The second may be sitting in somebody else's calendar: checking an answer, correcting a mistake, handling an escalation or explaining a result to a customer. If that work disappears from the budget, an apparently efficient system can quietly move effort rather than remove it. Capgemini's new snapshot surveyed 500 senior executives at organizations with more than $1 billion in annual revenue, across ten countries and thirteen industries, in September. Fifty-nine percent said AI spending had been significantly above expectations or budget. Among affected organizations, the reported average overrun was 24%. Those are respondents' assessments, not independently audited accounts, and the sample does not represent every small business or user. The release also says more than three in five executives expect a multi-model strategy within twelve months. Choosing a smaller or cheaper model for a bounded task can be sensible. But adding models can also add routing decisions, evaluations and operational complexity. A planned strategy is not a demonstrated saving. The hopeful part is that these are tractable management questions. Define an acceptable completed task, measure how often it succeeds without correction, and include the cost of people and controls needed to deliver it. Compare that result with a credible alternative, not merely with last month's token bill. We should resist two easy conclusions: that AI pays for itself because usage is growing, or that budget overruns prove AI has failed. Spending can rise because valuable work expands, because forecasts were poor, or because a workflow is inefficient. Distinguishing those explanations is the work the headline number cannot do for us.

3 min
Conceptual miniature film studio and separate counting apparatus, illustrating the difference between content tokens and computing resources.
Work & marketsChina+3 clusters03

China's viral AI-drama statistic does not measure national computing power

The arresting claim is that China spends more than half its AI computing power making short dramas. The company document behind the argument supports a narrower statement: COL's June investor record attributes 55% of domestic token consumption to short dramas and video generation. It does not provide a reproducible national accounting method. Recent reporting has revived the figure, while ByteDance has challenged the compute framing and described a different measure for its own operations. That distinction matters because a token count is not a standardized measure of hardware work across models and media. Nor does it establish electricity use, profit, scientific value or social benefit. A business can generate many tokens while creating little revenue; another can sell a valuable service with a small output. Comparing the two requires a consistent denominator and an explanation of what the units represent. There is a serious economic question underneath the viral statistic. Cheap synthetic video can change entry costs for small studios while making distribution, audience acquisition and repeated generation more important. Yet we should not turn a company's promotional account into a national verdict on useful versus wasteful AI. Entertainment is also work, income and culture. AIImpactLab's judgment is that the measurement problem belongs in the headline. The industry needs comparable reporting on workloads, spending and outcomes before policymakers or investors use this percentage to justify a different allocation of resources.

5 min
A mechanical confidence dial controls an answer gate while a separate correctness marker remains visibly misaligned.
Technical failuresGlobal+1 clusters04

Language models use internal confidence to decide when to abstain

A peer-reviewed study has moved the debate about AI uncertainty beyond asking whether a model can produce a confidence score. Across four language models, researchers used a four-phase experiment to test whether confidence-related internal states actually drive the decision to answer or abstain. Confidence strongly predicted refusal behavior. More importantly, activation steering that boosted or suppressed confidence changed abstention rates, and instructions that altered the decision threshold changed behavior without fundamentally changing the underlying confidence representation. That is causal evidence for a two-stage control process: an internal confidence signal and a policy that decides how much confidence is enough. The safety opportunity is real. Systems could be engineered to defer, verify, or request human review when their own uncertainty crosses a tested boundary. The warning is just as important. Verbal confidence independently influenced abstention even though it was less effective than calibrated token probabilities at distinguishing correct from incorrect answers. A model can therefore act on a confidence signal that is behaviorally powerful but imperfectly connected to truth. This is not evidence of consciousness, and the experiment does not show that open-ended agents can reliably monitor long reasoning chains. It used factual multiple-choice questions without chain-of-thought instructions. The practical lesson is narrower and more useful: confidence is a control surface. High-stakes deployment must validate both the internal signal and the threshold policy under real costs, because a model that knows when it feels unsure can still be confidently wrong about whether to proceed.

5 min
A red vulnerability trace crosses a technical model blueprint and exposes two fault points before meeting a transparent restricted-access gate.
SecurityGlobal+4 clusters05

Astra crossed OpenAI's critical cyber threshold before public release

OpenAI says its upcoming Astra model is the first of its systems to reach a critical cybersecurity capability threshold. With appropriate tools and access, the company says Astra can find previously unknown security flaws and develop exploit paths against well-protected systems without step-by-step human direction. Its internal evidence is striking: a perfect result on a known-vulnerability exploit benchmark, two zero-day flaws discovered in one exploit chain, a full browser-compromise chain that escaped a sandbox, and a local privilege-escalation path to root access. OpenAI says Astra is also more token-efficient than GPT-5.6 Sol in vulnerability discovery and exploit development. The safeguard results are material but not conclusive. OpenAI reports that Astra refused 91.5 percent of malicious cyber requests in a jailbreak evaluation, compared with 59 percent for GPT-5.6 Sol, and did not try to evade automated review in its tests. Advanced access will initially be restricted to trusted testers and defenders. Because the developer defines the category, controls the model, and benefits from release, critical capability claims and safety claims both need independent replication. Protected third-party testing, monitored access, zero-day disclosure, clear incident thresholds, and enforceable pause conditions should travel with the model wherever its access expands.

6 min