IBM Plans Claude Mythos in Concert Protect for Enterprises
IBM will build Anthropic's Claude Mythos into Concert Protect so customers can scan their own code, test exploitability and approve fixes. Launch is aimed at early 2027.
SUMO· the number puzzle
Dialo· learn languages by speaking
ediviewer· read EDI files like documents
giftdetect· the right gift in minutesIBM plans to build Anthropic's Claude Mythos into its Concert Protect security software, the company said on Monday. The aim is to let customers scan their own code for previously unknown flaws, with a target launch in early 2027.
Mythos would examine code that customers select and look for vulnerabilities that conventional scanners miss. Concert Protect would then test whether attackers could actually exploit each flaw in that customer's own systems. IBM and Red Hat tools would draft fixes, and developers must review and approve them before anything goes in. IBM cautioned that its plans could change.
Why exploitability is the bottleneck
"The bottleneck is no longer finding vulnerabilities. It is determining which can actually be exploited and fixing them before attackers do," IBM wrote. The gap is visible in Anthropic's restricted Project Glasswing programme, which flagged more than 10,000 possible serious flaws in its first month, while only 97 had been patched by then.
IBM says its earlier Mythos work through Glasswing was internal research, making this its first step toward selling the capability. That matches Anthropic's approach of placing Mythos inside partners' products, so users receive findings and fixes rather than the model itself. IBM is also not tied to a single lab: in June it joined OpenAI's cyber programme for a similar service.
Access stays tightly controlled
Mythos remains limited to vetted organisations because of how effective it is at finding security flaws. Anthropic briefly switched it off in June to comply with US export controls, and the White House now reportedly decides who joins its partner list. On Monday Anthropic released Mythos 5.2, which is also restricted to vetted organisations.
For enterprise security teams, the plan would offer a route to Mythos-level scanning without direct access to the model, bundled with validation and fix workflows that keep humans in the approval loop.
Apps worth a look
SUMO: the number puzzleTap the right tiles, add them up, hit the target. Easy to learn, hard to master. Free, for all ages.
Dialo: learn languages by speakingTalk to an AI conversation partner; every sentence comes back corrected and explained in your language. 9 languages, A1 to C2.
ediviewer: read EDI files like documentsOpen EDIFACT and X12 messages, see every segment explained, convert to JSON/XML, export PDF and Excel. Runs locally, no upload.
giftdetect: the right gift in minutesPick who it is for, the occasion and their personality. The search engine reads product descriptions from local shops and ranks what fits. Free, no sign-up.
More from this day
See all →
Solana weighs removing feature exploited in $270m DeFi hack
Vanderbilt University Hit by Suspected Ransomware Attack
Edgevana secures $700m in Nvidia compute deals for AI
