Anthropic says it cut the data rivals can siphon from Claude by up to 99% in just months.* The key isn't a single wall — it's three layers closing in from different angles.
AspireAI community · anthropic.com
*Anthropic says it cut the data rivals can siphon from Claude by up to 99% in just months.* The key isn't a single wall — it's three layers closing in from different angles.
The threats are industrial. Distillers run networks of tens of thousands of fake accounts asking Claude to reveal its hidden reasoning; one ring ran 20,000-plus at once. Anthropic now locks reasoning transcripts to the requesting organization (no more smuggling reasoning between accounts), runs a constitutional-classifier style two-stage filter with under 0.1% false positives, and bans repeat offenders — charging for blocked requests so attempts aren't free. Reasoning tokens exposed on Anthropic's own platform fell 10-100x; on some days, too small to measure.
Anthropic names Z.ai's GLM-5.3 and a Kimi request affiliated with China's PLA as examples. The honest framing still stands — vendors say this is "cat-and-mouse" and they haven't seen a sophisticated persistent attacker yet.
Read the story and explore more from the AspireAI community.
Two critical PHP remote code execution flaws — one found by GPT-5.6 Sol — were patched on October 21, 2026, and OpenAI published a disclosure framework alongside the fix. Fixed versions: PHP 8.5.12, 8.4.27, 8.3.36 and 8.2.35. Both allow unauthenticated remote code execution, and PHP's advisory says risk is highest for mod_php, where a demonstrated exploit triggers before any application code runs. The CVEs are distinct failure modes. CVE-2026-109618 (OOM corruption): out-of-memory handling can interrupt PHP while internal request data is only partially initialized, and request cleanup may then treat the invalid data as valid. CVE-2026-109619 (stale global state): request-owned internal state can survive after its backing memory is freed, and a later request served by the same worker may use that stale state. OpenAI states GPT-5.6 Sol identified the first flaw through its Patch the Planet initiative, and that Wiz independently discovered and reported both vulnerabilities. PHP's maintainers led the fixes and release, with Wiz, Cloudflare and OpenAI contributing research and coordination. Wiz's write-up adds that it used GPT-5.6 Sol and Gemini 4 Argon throughout, with Gemini 4 Argon contributing to finding the vulnerability and validating the patch. OpenAI describes the gap between discovery and public exploitation as a "Defender's Window," proposes a five-stage response framework, and committed $1 billion in subsidised Daybreak model access for organizations with limited security resources. Honest caveats: these are OpenAI's and Wiz's own characterizations, and OpenAI also notes the same capabilities help attackers build offensive tooling faster. PHP 8.1 and older no longer receive updates from the PHP project.