Quick take: This morning’s strongest developments are less about spectacle than accountability: a failed drug trial that must still yield knowledge, a court remedy aimed at the hidden auction behind much of the web, a clearer measure of polar ice loss, and a European attempt to close the climate-insurance gap.
An ALS drug fails its decisive test
Novartis confirmed on 16 September that it had stopped developing lifonebart (VHB937) for amyotrophic lateral sclerosis after a 251-person mid-stage trial failed both its primary and secondary endpoints. The drug was designed to stimulate TREM2, a receptor involved in clearing damaged cells and regulating inflammation. A negative result is painful for people living with ALS, but it is also useful evidence: it narrows the field and can prevent patients and researchers investing further hope in an unsupported mechanism. The limitation is that the company’s Alzheimer’s study of the same drug continues; success or failure there cannot be inferred from the ALS result because the diseases, populations and endpoints differ. Individuals considering experimental treatment should distinguish biological plausibility from demonstrated benefit and ask whether complete results—not only a company summary—will be released. Researchers, funders and patient groups can press for prompt registry updates and sufficiently detailed negative data so that this trial reduces duplication rather than simply disappearing. Source: Reuters, 16 September 2026.
Possible Alkemata article: When a Failed Trial Is Still Useful Knowledge
A judge changes the machinery of Google’s ad auctions
A US federal judge ordered Google on 16 September to loosen the links between its publisher ad server and AdX exchange, modify auction rules and install an internal antitrust compliance officer for six years. The remedy follows an earlier finding that Google monopolised parts of advertising technology; it stops short of the structural breakup sought by the government, and Google says it will appeal. The promising part is practical rather than symbolic: publishers should gain more freedom to use competing exchanges, potentially reducing the toll taken by one vertically integrated intermediary. Yet implementation details matter. Technical access can exist on paper while switching costs, data advantages or auction design preserve the old imbalance. Readers may feel the consequences indirectly through the finances and independence of ad-funded news and specialist sites. Publishers should document fees, latency and auction outcomes before and after the changes, while advertisers can demand intelligible supply-chain reporting. The useful question is not whether Google “won” or “lost”, but whether competing routes genuinely become viable. Sources: Reuters and the US Department of Justice, 16 September 2026.
Possible Alkemata article: Who Controls the Auction Behind the Web?
Forty-four years of observations sharpen the polar-ice picture
A study published on 16 September combined 42 independent estimates from 27 satellite missions to reconstruct Greenland and Antarctic ice-sheet mass from 1979 to 2023. The result is unusually comprehensive: the two ice sheets lost an estimated 11,309 ± 565 billion tonnes, adding 31.4 ± 1.6 millimetres to global mean sea level. About 84% of the loss came from faster glacier flow and ice discharge, with the remainder from reduced surface mass balance. Bringing separate measurement methods into one reconciled record gives planners a firmer baseline and makes disagreements visible. It does not make local sea-level projections simple: land movement, currents and coastal engineering still alter what any town experiences. Nor does the slower loss observed in 2020–23 establish a reversal; the authors attribute it to short-term snowfall and summer conditions within a much longer trend. Individuals in coastal areas can check whether municipal flood maps and infrastructure plans use current sea-level assumptions, rather than treating a global average as a local forecast. Source: Scientific Data, 16 September 2026; context: Reuters.
Possible Alkemata article: What Three Centimetres of Sea-Level Rise Actually Changes
Europe proposes a climate-insurance alliance—but not yet a safety net
The European Commission proposed on 16 September a “climate insurance alliance” bringing together insurers, investors, public authorities, risk modellers and buyers after another summer of extreme weather. The starting problem is stark: only about a quarter of Europe’s catastrophic losses are privately insured. Shared or parametric products—paying when an agreed trigger is reached—could make recovery faster and help cover risks that conventional policies increasingly avoid. But this remains an initiative, not an insurance product or guarantee, and a trigger can fail people whose losses are real but fall just outside its definition. There is also a political danger in using financial products as a substitute for flood protection, heat planning, building standards or public solidarity. Households and small businesses can act now by asking, in writing, whether their current policies cover flood, wildfire and business interruption, what exclusions apply, and how payouts are calculated. Citizens should also ask governments which risks will remain collectively borne: climate adaptation cannot be reduced to deciding who can still afford a premium. Source: Reuters, 16 September 2026.
Possible Alkemata article: Insurance Cannot Replace Climate Adaptation
