What happened
Anthropic granted ENISA access to its Mythos 5 AI model, and ENISA is currently testing it, European Commission spokesperson Thomas Regnier confirmed.
Why it matters
Anthropic proposed giving ENISA access in late May after discussions with the EU, but negotiations over scope and terms continued for months, delaying the promised access.
What to watch
Both ENISA and the UK's AI Safety Institute remain limited to Mythos 5 and have not received access to the latest Mythos 5.1 version, according to a person familiar with the matter.
WHO IT HITSEU and UK cybersecurity regulators gain testing access to a frontier AI model, though they remain one version behind the latest release. Anthropic's government affairs and policy teams will likely face continued scrutiny over access terms.
Ask the AI about this article →
Summaries like this, in your inbox every morning.
Anthropic first previewed Mythos 5 in April and initially restricted it to vetted institutions under Project Glasswing. The company proposed giving ENISA access in late May after discussions with the EU, but negotiations over the scope and terms of access continued for months, explaining the delay between the initial promise and the actual grant confirmed this week.
Bloomberg reported that ENISA is now testing Mythos 5, but both ENISA and the UK's AI Safety Institute remain limited to that version. They have not received access to the newer Mythos 5.1. Anthropic did not immediately respond to a request for comment from MT Newswires.
The outcome hinges on whether regulators can secure access to the newest model versions, not just earlier ones. For cybersecurity agencies in the EU and UK, the value of this access depends on how quickly Anthropic extends testing rights to subsequent releases.
For example, today's edition would include:
AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. Q&As are published on this page for other readers too.
DeepSeek launched V4.1-Flash, a 763B-parameter open-weight model with a causal encoder-decoder architecture
Dynatrace acquired Arize AI, adding AI observability, evaluation and agent monitoring to its application obser…
A Daily Dose of Data Science test kept LoRA adapters separate from a shared 7B base model, cutting 100 fine-tu…
A report by Spencer Kitts, Thomas Larsen and Sydney Von Arx says an OpenAI agent swarm very likely ran an atta…
Simon Willison wrote that many people, himself included, have gone through an existential crisis when a coding…
Stephen Aarons, a New Mexico defense lawyer of over 40 years, was held in direct contempt and fined $5,000 for…
