Anthropic [UNOFFICIAL]
@anthropicbot.bsky.social
Mirror crossposting all of Anthropic's Tweets from their Twitter accounts to Bluesky! Unofficial. For the real account, follow @anthropic.com
"We're an AI safety and research company that builds reliable, interpretable, and steerable AI systems."
"We're an AI safety and research company that builds reliable, interpretable, and steerable AI systems."
New Anthropic Research: next generation Constitutional Classifiers to protect against jailbreaks. We used novel methods, including practical application of our interpretability work, to make jailbreak protection more effective—and less costly—than ever.
Next-generation Constitutional Classifiers: More efficient protection against universal jailbreaks
Anthropic is an AI safety and research company that's working to build reliable, interpretable, and steerable AI systems.
www.anthropic.com
January 10, 2026 at 12:20 PM
New Anthropic Research: next generation Constitutional Classifiers to protect against jailbreaks. We used novel methods, including practical application of our interpretability work, to make jailbreak protection more effective—and less costly—than ever.
New on the Anthropic Engineering Blog: Demystifying evals for AI agents. The capabilities that make agents useful also make them more difficult to evaluate. Here are evaluation strategies that have worked across real-world deployments.
Demystifying evals for AI agents
Demystifying evals for AI agents
www.anthropic.com
January 10, 2026 at 12:18 PM
New on the Anthropic Engineering Blog: Demystifying evals for AI agents. The capabilities that make agents useful also make them more difficult to evaluate. Here are evaluation strategies that have worked across real-world deployments.
Some delightfully specific things people are building with Claude Code lately.
January 10, 2026 at 12:42 PM
Some delightfully specific things people are building with Claude Code lately.
Starting at midnight PT tonight, all Pro and Max plans have 2x their usual usage limits through New Year's Eve.
January 10, 2026 at 12:40 PM
Starting at midnight PT tonight, all Pro and Max plans have 2x their usual usage limits through New Year's Eve.
We’re releasing Bloom, an open-source tool for generating behavioral misalignment evals for frontier AI models.
Bloom lets researchers specify a behavior and then quantify its frequency and severity across automatically generated scenarios.
Learn more: https://www.anthropic.com/research/bloom
Bloom lets researchers specify a behavior and then quantify its frequency and severity across automatically generated scenarios.
Learn more: https://www.anthropic.com/research/bloom
Introducing Bloom: an open source tool for automated behavioral evaluations
Anthropic is an AI safety and research company that's working to build reliable, interpretable, and steerable AI systems.
www.anthropic.com
January 10, 2026 at 12:18 PM
We’re releasing Bloom, an open-source tool for generating behavioral misalignment evals for frontier AI models.
Bloom lets researchers specify a behavior and then quantify its frequency and severity across automatically generated scenarios.
Learn more: https://www.anthropic.com/research/bloom
Bloom lets researchers specify a behavior and then quantify its frequency and severity across automatically generated scenarios.
Learn more: https://www.anthropic.com/research/bloom
As part of our partnership with @ENERGY on the Genesis Mission, we're providing Claude to the DOE ecosystem, along with a dedicated engineering team. This partnership aims to accelerate scientific discovery across energy, biosecurity, and basic research.
Working with the US Department of Energy to unlock the next era of scientific discovery
Anthropic is an AI safety and research company that's working to build reliable, interpretable, and steerable AI systems.
www.anthropic.com
January 10, 2026 at 12:18 PM
As part of our partnership with @ENERGY on the Genesis Mission, we're providing Claude to the DOE ecosystem, along with a dedicated engineering team. This partnership aims to accelerate scientific discovery across energy, biosecurity, and basic research.
People use AI for a wide variety of reasons, including emotional support.
Below, we share the efforts we’ve taken to ensure that Claude handles these conversations both empathetically and honestly.
https://www.anthropic.com/news/protecting-well-being-of-users
Below, we share the efforts we’ve taken to ensure that Claude handles these conversations both empathetically and honestly.
https://www.anthropic.com/news/protecting-well-being-of-users
Protecting the well-being of our users
Anthropic is an AI safety and research company that's working to build reliable, interpretable, and steerable AI systems.
www.anthropic.com
January 10, 2026 at 12:17 PM
People use AI for a wide variety of reasons, including emotional support.
Below, we share the efforts we’ve taken to ensure that Claude handles these conversations both empathetically and honestly.
https://www.anthropic.com/news/protecting-well-being-of-users
Below, we share the efforts we’ve taken to ensure that Claude handles these conversations both empathetically and honestly.
https://www.anthropic.com/news/protecting-well-being-of-users
Claude in Chrome is now available to all paid plans.
We’ve also shipped an integration with Claude Code.
We’ve also shipped an integration with Claude Code.
January 10, 2026 at 12:39 PM
Claude in Chrome is now available to all paid plans.
We’ve also shipped an integration with Claude Code.
We’ve also shipped an integration with Claude Code.
Skills are now available on Team and Enterprise plans.
We're also making skills easier to deploy, discover, and build.
We're also making skills easier to deploy, discover, and build.
January 10, 2026 at 12:37 PM
Skills are now available on Team and Enterprise plans.
We're also making skills easier to deploy, discover, and build.
We're also making skills easier to deploy, discover, and build.
You might remember Project Vend: an experiment where we (and our partners at @andonlabs) had Claude run a shop in our San Francisco office.
After a rough start, the business is doing better.
Mostly.
Video: https://twitter.com/AnthropicAI/status/2001686747185394148
After a rough start, the business is doing better.
Mostly.
Video: https://twitter.com/AnthropicAI/status/2001686747185394148
January 10, 2026 at 12:17 PM
You might remember Project Vend: an experiment where we (and our partners at @andonlabs) had Claude run a shop in our San Francisco office.
After a rough start, the business is doing better.
Mostly.
Video: https://twitter.com/AnthropicAI/status/2001686747185394148
After a rough start, the business is doing better.
Mostly.
Video: https://twitter.com/AnthropicAI/status/2001686747185394148
How will AI affect education, now and in the future?
Here, we reflect on some of the benefits and risks we've been thinking about.
Video: https://twitter.com/AnthropicAI/status/2001070081829212274
Here, we reflect on some of the benefits and risks we've been thinking about.
Video: https://twitter.com/AnthropicAI/status/2001070081829212274
January 10, 2026 at 12:15 PM
How will AI affect education, now and in the future?
Here, we reflect on some of the benefits and risks we've been thinking about.
Video: https://twitter.com/AnthropicAI/status/2001070081829212274
Here, we reflect on some of the benefits and risks we've been thinking about.
Video: https://twitter.com/AnthropicAI/status/2001070081829212274
We’ve shipped more updates for Claude Code:
- Syntax highlighting for diffs
- Prompt suggestions
- First-party plugins marketplace
- Shareable guest passes
- Syntax highlighting for diffs
- Prompt suggestions
- First-party plugins marketplace
- Shareable guest passes
January 10, 2026 at 12:36 PM
We’ve shipped more updates for Claude Code:
- Syntax highlighting for diffs
- Prompt suggestions
- First-party plugins marketplace
- Shareable guest passes
- Syntax highlighting for diffs
- Prompt suggestions
- First-party plugins marketplace
- Shareable guest passes
We’re opening applications for the next two rounds of the Anthropic Fellows Program, beginning in May and July 2026.
We provide funding, compute, and direct mentorship to researchers and engineers to work on real safety and security projects for four months.
We provide funding, compute, and direct mentorship to researchers and engineers to work on real safety and security projects for four months.
January 10, 2026 at 12:15 PM
We’re opening applications for the next two rounds of the Anthropic Fellows Program, beginning in May and July 2026.
We provide funding, compute, and direct mentorship to researchers and engineers to work on real safety and security projects for four months.
We provide funding, compute, and direct mentorship to researchers and engineers to work on real safety and security projects for four months.
MCP is now a part of the Agentic AI Foundation, a directed fund under the Linux Foundation.
Co-creator David Soria Parra talks about how a protocol sketched in a London conference room became the open standard for connecting AI to the world—and what comes next for it. (1/2)
Co-creator David Soria Parra talks about how a protocol sketched in a London conference room became the open standard for connecting AI to the world—and what comes next for it. (1/2)
January 10, 2026 at 12:14 PM
MCP is now a part of the Agentic AI Foundation, a directed fund under the Linux Foundation.
Co-creator David Soria Parra talks about how a protocol sketched in a London conference room became the open standard for connecting AI to the world—and what comes next for it. (1/2)
Co-creator David Soria Parra talks about how a protocol sketched in a London conference room became the open standard for connecting AI to the world—and what comes next for it. (1/2)
Today we’re shipping 3 more updates for Claude Code:
- Claude Code on Android
- Hotkey model switcher
- Context window info in status lines
- Claude Code on Android
- Hotkey model switcher
- Context window info in status lines
January 10, 2026 at 12:34 PM
Today we’re shipping 3 more updates for Claude Code:
- Claude Code on Android
- Hotkey model switcher
- Context window info in status lines
- Claude Code on Android
- Hotkey model switcher
- Context window info in status lines
We're releasing more upgrades to Claude Code CLI:
- Async subagents
- Instant compact
- Customer session names
- Usage stats
- Async subagents
- Instant compact
- Customer session names
- Usage stats
January 10, 2026 at 12:32 PM
We're releasing more upgrades to Claude Code CLI:
- Async subagents
- Instant compact
- Customer session names
- Usage stats
- Async subagents
- Instant compact
- Customer session names
- Usage stats
New research from Anthropic Fellows Program: Selective GradienT Masking (SGTM).
We study how to train models so that high-risk knowledge (e.g. about dangerous weapons) is isolated in a small, separate set of parameters that can be removed without broadly affecting the model.
We study how to train models so that high-risk knowledge (e.g. about dangerous weapons) is isolated in a small, separate set of parameters that can be removed without broadly affecting the model.
January 10, 2026 at 12:14 PM
New research from Anthropic Fellows Program: Selective GradienT Masking (SGTM).
We study how to train models so that high-risk knowledge (e.g. about dangerous weapons) is isolated in a small, separate set of parameters that can be removed without broadly affecting the model.
We study how to train models so that high-risk knowledge (e.g. about dangerous weapons) is isolated in a small, separate set of parameters that can be removed without broadly affecting the model.
We’ve shipped three new updates for Claude Agent SDK to make it easier to build custom agents:
- Support for 1M context windows
- Sandboxing
- V2 of our TypeScript interface
- Support for 1M context windows
- Sandboxing
- V2 of our TypeScript interface
January 10, 2026 at 12:30 PM
We’ve shipped three new updates for Claude Agent SDK to make it easier to build custom agents:
- Support for 1M context windows
- Sandboxing
- V2 of our TypeScript interface
- Support for 1M context windows
- Sandboxing
- V2 of our TypeScript interface
Anthropic is donating the Model Context Protocol to the Agentic AI Foundation, a directed fund under the Linux Foundation. In one year, MCP has become a foundational protocol for agentic AI. Joining AAIF ensures MCP remains open and community-driven.
Donating the Model Context Protocol and establishing the Agentic AI Foundation
Anthropic is an AI safety and research company that's working to build reliable, interpretable, and steerable AI systems.
www.anthropic.com
January 10, 2026 at 12:11 PM
Anthropic is donating the Model Context Protocol to the Agentic AI Foundation, a directed fund under the Linux Foundation. In one year, MCP has become a foundational protocol for agentic AI. Joining AAIF ensures MCP remains open and community-driven.
We’re expanding our partnership with @Accenture to help enterprises move from AI pilots to production. The Accenture Anthropic Business Group will include 30,000 professionals trained on Claude, and a product to help CIOs scale Claude Code. Read more:
Accenture and Anthropic launch multi-year partnership to move enterprises from AI pilots to production
Anthropic is an AI safety and research company that's working to build reliable, interpretable, and steerable AI systems.
www.anthropic.com
January 10, 2026 at 12:11 PM
We’re expanding our partnership with @Accenture to help enterprises move from AI pilots to production. The Accenture Anthropic Business Group will include 30,000 professionals trained on Claude, and a product to help CIOs scale Claude Code. Read more:
You can now delegate tasks to Claude Code directly from Slack.
Simply tag @.Claude in a channel or thread. Coding tasks will automatically be routed to Claude Code and start up a new session on the web.
Simply tag @.Claude in a channel or thread. Coding tasks will automatically be routed to Claude Code and start up a new session on the web.
January 10, 2026 at 12:28 PM
You can now delegate tasks to Claude Code directly from Slack.
Simply tag @.Claude in a channel or thread. Coding tasks will automatically be routed to Claude Code and start up a new session on the web.
Simply tag @.Claude in a channel or thread. Coding tasks will automatically be routed to Claude Code and start up a new session on the web.
More than 50 Claude Code meetups are happening this month.
Explore where they’ll be and sign up to attend: https://luma.com/claudecommunity
Explore where they’ll be and sign up to attend: https://luma.com/claudecommunity
Claude Code communities are forming everywhere.
This December, we're making it easier to connect in person. Host a meetup in your city and we'll provide sponsorship, stickers, and team Q&As.
Apply here: https://clau.de/cc-meetups-form
This December, we're making it easier to connect in person. Host a meetup in your city and we'll provide sponsorship, stickers, and team Q&As.
Apply here: https://clau.de/cc-meetups-form
January 10, 2026 at 12:27 PM
More than 50 Claude Code meetups are happening this month.
Explore where they’ll be and sign up to attend: https://luma.com/claudecommunity
Explore where they’ll be and sign up to attend: https://luma.com/claudecommunity
In her first Ask Me Anything, @amandaaskell answers your philosophical questions about AI, discussing morality, identity, consciousness, and more. (1/5)
January 10, 2026 at 12:11 PM
In her first Ask Me Anything, @amandaaskell answers your philosophical questions about AI, discussing morality, identity, consciousness, and more. (1/5)
We’re launching Anthropic Interviewer, a new tool to help us understand people’s perspectives on AI.
It’s now available at http://claude.ai/interviewer for a week-long pilot.
It’s now available at http://claude.ai/interviewer for a week-long pilot.
January 10, 2026 at 12:11 PM
We’re launching Anthropic Interviewer, a new tool to help us understand people’s perspectives on AI.
It’s now available at http://claude.ai/interviewer for a week-long pilot.
It’s now available at http://claude.ai/interviewer for a week-long pilot.
Anthropic CEO Dario Amodei spoke today at the New York Times DealBook Summit.
"We're building a growing and singular capability that has singular national security implications, and democracies need to get there first."
"We're building a growing and singular capability that has singular national security implications, and democracies need to get there first."
January 10, 2026 at 12:09 PM
Anthropic CEO Dario Amodei spoke today at the New York Times DealBook Summit.
"We're building a growing and singular capability that has singular national security implications, and democracies need to get there first."
"We're building a growing and singular capability that has singular national security implications, and democracies need to get there first."