Tooling4d ago
X discussion: viral posts claim Anthropic "quietly killed" Claude's reverse engineering with the REA toolkit after a Samsung TV firmware teardown went viral; no official statement yet
One of the hottest AI developer topics on X over the past day. At 21:55 UTC on 2026-10-23, security account @0x0SojalSec (Md Ismail Šojal) posted that "Claude can no longer Reverse Engineer Anything" and that "Anthropic just quietly killed REA", tying the timing to viral cases of people using Claude to reverse engineer Samsung TV firmware and find data-collection services; the post had about 8,345 likes, 1,055 reposts and 514K views. An earlier post making the same claim by @AbhinavXJ had about 10,477 likes and 640K views. As of this run we found no statement from Anthropic or its staff, and some users replied that it still works for them. These are user reports and opinions on X, not a confirmed policy change.
Industry20d ago
Anthropic expands its Cyber Verification Program into three access tiers with Opus 5.5, Sonnet 5.5 and Mythos 5.1; Red Team tier completes 34/50 CyScenarioBench tasks with zero blocks
On 2026-10-06 Anthropic merged Project Glasswing and the original Cyber Verification Program into one three-tier program: Defense Access (SOC, incident response, malware reversing, vuln validation), Red Team Access (authorized pentesting) and Specialized Access (safety-critical systems such as power grids, flight systems and telecom, vetted with the US government). All tiers get Claude Opus 5.5, Sonnet 5.5 and Mythos 5.1 with reduced cyber blocking. On CyScenarioBench, GA models were blocked on the first prompt of every task, Defense tier blocked 46/50 trials, and Red Team tier had zero blocks and completed 34/50 (matching the 67.6% no-safeguard rate). Glasswing partners reported at least 129,000 verified vulnerabilities from April to July 2026.
Research10d ago
Claude traces the NIST elliptic-curve seeds to lrand48 seeded with 1997–98 timestamps, settling a 25-year backdoor question
Anthropic says an internal research model recovered, in under half an hour, how the seeds of NIST P-192/224/256/384/521 were produced: lrand48 output seeded with Unix timestamps from 1997–98, not a hashed phrase. Bounty organizer Filippo Valsorda independently reproduced it in Go; the $36,864 bounty goes to ISRG. This is not a break and no action is needed.
Discussion & Comments
0Sign in to join the discussion
Connect with AI developers to exchange benchmark insights.