Tag

Fable 5

All articles tagged with #fable 5

technology8 days ago•72 min saved

Opus 5: Cheaper AI coding power raises new security and budgeting questions

Anthropic's Opus 5 delivers state-of-the-art coding performance for just $5 per million input tokens and $25 per million output tokens, outperforming Fable 5 on benchmarks and enabling longer autonomous tasks. However, the lower running cost intensifies concerns about security, spend management, and oversight for agentic AI, prompting a shift toward short-lived credentials, microVM isolation, and smarter telemetry; operators may route unsafe tasks to Opus 4.8 via Automatic Fallbacks, while balancing biology tasks and long-horizon workflows.

via The New Stack|

#anthropic #fable-5 #opus-5

technology8 days ago•3 min saved

Anthropic debuts Claude Opus 5, near-Fable intelligence at half the cost

Anthropic released Claude Opus 5, a more efficient and capable model that approaches the capabilities of Fable 5 at half the price, becoming the default model on Claude Max and the strongest option on Claude Pro. The release also adds mid-conversation tool changes and automatic API fallbacks to route requests to the best available model when needed. Opus 5 is priced at $5 per million input tokens and $25 per million output tokens, and is available today across platforms.

via 9to5Mac|

#anthropic #claude #fable-5

technology9 days ago•2 min saved

Anthropic rolls out Opus 5, aiming for Fable 5 parity with stronger safeguards

Anthropic released Claude Opus 5, a model pitched as close to Claude Fable 5 in many domains and stronger at coding, with enhanced cybersecurity safeguards versus Opus 4.8. Opus 5 is priced at $5 input / $25 output per million tokens, cheaper than Fable 5 and GPT-5.6, and offers a Fast mode for higher speed plus an automatic fallback to a lower-tier model when safeguards are triggered, targeting enterprise use and addressing regulatory scrutiny.

via The Verge|

#ai-regulation #anthropic #cybersecurity-safeguards

technology13 days ago•5 min saved

AI tool Fable 5 allegedly disproves the Jacobian conjecture

A mathematician, Levent Alpöge, used Anthropic’s Fable 5 to claim a counterexample to the Jacobian conjecture, a nearly 90-year open problem in algebraic geometry. Experts warn that a single counterexample, while notable, is not equivalent to a full proof and may offer limited insight, though the episode underscores AI’s potential in assisting mathematical research and discovery, with discussions also touching on OpenAI’s related progress.

via Mashable|

#academia #artificial-intelligence #fable-5

science14 days ago•4 min saved

AI-Engine Counterexample Cracks Century-Old Jacobian Conjecture

Mathematician Levent Alpöge used Anthropic’s Claude Fable 5 to forge a concrete three-variable polynomial with Jacobian determinant -2 that is not invertible, producing a verifiable counterexample to the Jacobian conjecture (pending arXiv peer review); the episode underscores AI as a genuine research collaborator and highlights the rise of formal verification like Lean in AI-driven mathematics.

via Glitchwire|

#ai-math #anthropic #arxiv

technology14 days ago•9 min saved

Alibaba's Qwen 3.8: open-weight AI aims for frontier parity behind Fable 5

Alibaba unveils Qwen 3.8, a 2.4-trillion-parameter open-weight AI model that the company says can match frontier models and sit just behind Fable 5. It supports multimodal inputs (images, videos, documents), promises improved coding and productivity tasks over Qwen 3.7-Max, and will offer open weights soon with a preview of Qoder at 10% of standard pricing.

via the-decoder.com|

#alibaba #fable-5 #frontier-models

technology1 month ago•9 min saved

Anthropic’s Claude Fable 5 and Mythos 5 Go Global After Safety Push

The U.S. has lifted export restrictions on Anthropic’s Claude models, with Fable 5 going global and Mythos 5 access restored for US users, following government testing and safeguards. Anthropic expanded its government partnerships, opened red-teaming, introduced a 24/7 monitoring team and a HackerOne program to surface jailbreaks, and plans to broaden Mythos access through the Glasswing program, while warning that export controls could be reinstated.

via Ars Technica|

#ai-safety #anthropic #export-controls

technology1 month ago•3 min saved

US Lifts Export Curbs, Restoring Access to Anthropic’s Fable 5 and Mythos 5

The U.S. Department of Commerce removed export controls on Anthropic’s Claude-based Fable 5 and Mythos 5, allowing access for U.S. organizations that operate and defend critical infrastructure and paving the way for general public access after Anthropic agreed to enhanced security measures and ongoing government collaboration on standards and monitoring for malicious activity.

via Al Jazeera|

#anthropic #export-controls #fable-5

technology1 month ago•1 min saved

Anthropic clears Mythos 5 redeployment to select US critical-infrastructure groups

Anthropic says the US government has authorized redeploying Mythos 5 to a set of US organizations that defend critical infrastructure, after a directive earlier this month banned foreign nationals from using the firm’s Mythos 5 and Fable 5 models. The company will restore access to Mythos 5 quickly for these groups and continues working with the government to expand access and bring Fable 5 back for general use, as EU officials express concern about dependence on foreign technologies.

via Euronews.com|

#anthropic #critical-infrastructure #fable-5

technology1 month ago•4 min saved

US Nears Reopening Anthropic's Fable 5 After Security Hold

The Trump administration is close to restoring access to Anthropic's Fable 5 after a 15‑day outage over security concerns, with limited Mythos 5 access already permitted by the Commerce Department; final approvals from the Pentagon and NSA are still pending, but the move signals a thaw in a months‑long dispute and underscores a push for a formal, transparent process to vet and regulate powerful AI models.

via Axios|

#ai-policy #anthropic #fable-5

technology1 month ago•1 min saved

US Clears Path for Anthropic Mythos 5 Access by Trusted Institutions

The U.S. Commerce Department lifted export controls on Anthropic’s Claude Mythos 5, allowing release to more than 100 trusted US institutions as safeguards are deemed sufficient; a future release of Fable 5 remains under discussion. The move coincides with OpenAI’s GPT-5.6 rollout to government-approved partners, and under the new terms licenses are no longer required to export Mythos 5 to Annex A entities or Anthropic’s foreign national employees.

via Semafor|

#anthropic #export-controls #fable-5

technology1 month ago•7 min saved

Mythos Allegedly Breaches NSA, Triggering Global AI Export Controls

An unverified claim that Anthropic’s Mythos hacked NSA classified systems in hours during an authorized red-team exercise sparked an export‑control ban that shut down Mythos and Fable 5 for non‑US users; it underscores gaps in frontier AI governance, affects Five Eyes partners, and leaves the models offline as policy, verification, and sovereignty questions unfold.

via TechSpot|

#anthropic #export-controls #fable-5

technology1 month ago•31 min saved

Regulatory Crossfire: How a 90-Minute Ultimatum Shook Anthropic’s Fable and the AI Industry

Verge’s Decoder episode unpacks how US export controls forced Anthropic to take down Fable 5 and Mythos 5, triggering a weekend scramble and a broader debate about AI safety and regulation. It traces the sequence from Amazon‑reported jailbreak concerns to a 90‑minute ultimatum and a regulatory approach that could either be more rigorous or more haphazard, highlighting tensions between government oversight, industry risk, and the push for a non‑voluntary framework while noting potential rollback of foreign‑nation access.

via The Verge|

#ai-regulation #anthropic #export-controls

technology1 month ago•1 min saved

Kalshi traders expect quick return of Anthropic's Fable 5 for U.S. users

Traders on Kalshi assign a 58% chance Anthropic will restore access to Fable 5 for U.S. customers by July 1, rising to 74% by July 10, after a U.S. government order suspended access; Anthropic has met with federal officials to discuss the issue, and Polymarket shows about 67% odds of restoration by July 1.

via CNBC|

#anthropic #fable-5 #kalshi

artificial-intelligence1 month ago•15 min saved

White House Readied Export Controls Against Claude Mythos Weeks Before Fable 5 Release

The Washington Post reports the White House had an export-control directive prepared to crack down on Anthropic’s Claude Mythos well before Claude Fable 5 went offline, after Anthropic expanded Project Glasswing access to about 111 entities (with roughly 50 already granted) including a South Korean telecom linked to China. An Amazon tip from Andy Jassy about jailbreak capabilities appears to have pushed policymakers toward action, and Anthropic later revoked access to the sensitive organization, contributing to Fable 5’s shutdown and underscoring a tense clash between frontier AI access and export controls.

via Gizmodo|

#anthropic #artificial-intelligence #claude-mythos