Answers from the documentation; checking which model is on duty
Hi - I answer from the OpenSmartRoute documentation: routing, the API, plans and quotas, self-hosting. Ask away, or open a support ticket if you need a person.
Grounded in the docs - follow a source before acting on it.
UK Launches AI Biosecurity Lab with OpenAI and Anthropic Funding - OpenSmartRoute
UK Launches AI Biosecurity Lab with OpenAI and Anthropic Funding
The UK government created the AI Biosecurity Lab with £100 million funding and £200 million from OpenAI and Anthropic. This lab aims to use frontier AI to detect and respond to biological threats.
Key points
The UK government announced the AI Biosecurity Lab on October 24, 2026.
OpenAI will provide £100 million in compute and model credits.
Anthropic will provide up to £100 million in model credits.
The lab will use GPT-Rosalind and Gemini models for research.
Why it matters: Safety teams can now access trusted frontier models for biological threat detection.
By OpenSmartRoute editorial · written through the router by writer-small
From Unite.AI - “UK Establishes AI Biosecurity Lab Backed by OpenAI and Anthropic”
The UK government announced a new AI Biosecurity Lab on October 24, 2026. This lab combines £100 million in public money with £200 million in private commitments. OpenAI and Anthropic are leading the private funding effort. The Cabinet Office released this news officially.
This lab aims to use advanced artificial intelligence to fight biological threats. Experts say AI can help discover new medicines faster. But they also warn that AI could lower barriers to making harmful biological agents. The government wants to balance these risks carefully.
The lab is the first dedicated organization of its kind in the world. It builds on the UK's existing Biological Security Strategy. The overall package involves up to £300 million in total funding. This includes both government grants and company pledges.
The lab will focus on early detection of new diseases. It will also speed up vaccine and treatment development. The goal is to help the UK respond faster to outbreaks. This includes strengthening resilience against future biological risks.
Funding Details and Company Commitments
OpenAI and Anthropic have committed up to £200 million to support this lab. OpenAI will provide £100 million in compute and model credits. These credits allow researchers to run large AI systems without buying hardware. Anthropic will also provide £100 million in model credits.
Anthropic will offer trusted model access and dedicated technical support. This means experts can use their specific AI tools safely. The lab will launch with these financial resources immediately.
OpenAI is providing access to its Rosalind Biodefense programme. This programme sponsors trusted developers building defensive AI applications. The focus is on life sciences research using specific models. Anthropic will work on similar technical support and access.
Research Partnerships with DeepMind and Basecamp
The lab will partner with Google DeepMind and Basecamp Research. DeepMind scientists and biosecurity experts will work closely with the lab. They will provide trusted access to frontier Gemini models. These models are designed for complex scientific tasks.
Basecamp Research is a UK-based AI life sciences company. They will provide access to their AI-biology models. These models tackle priority biosecurity challenges directly. The lab will use these tools to solve real problems.
Anthropic lets outside evaluators work inside its offices to test model safety. The move follows a pledge to improve oversight amid growing concerns about powerful AI systems.
Partnerships with public-sector organisations are central to the lab's approach. Academic institutions and industry partners will join the effort. The UK life sciences industry will also benefit from this work. International partners may join the collaboration network.
The Rosalind Biodefense Programme
OpenAI launched the Rosalind Biodefense programme on May 29, 2026. This programme supports work in epidemiological modeling and early detection. It focuses on non-pharmaceutical interventions as well. The programme expands trusted access to GPT-Rosalind for select partners.
GPT-Rosalind is a frontier reasoning model for life sciences research. It helps with screening, preparedness, and disease analysis. The programme supports work across multiple critical areas. OpenAI named several organisations as initial supporters of this programme.
Fourth Eon Biosecurity is one of the first supported organisations. Lawrence Livermore National Laboratory is another key partner. Johns Hopkins Applied Physics Laboratory joins this initial group. The Coalition for Epidemic Preparedness Innovations is also a supporter.
Independent Structure and National Initiative
The AI Biosecurity Lab will operate as an independent, non-profit organisation. It is designed to establish the UK as a leader in AI biosecurity. The structure allows it to move with the speed of a frontier company.
The lab operates within a National AI Biosecurity Initiative. This initiative brings together several government agencies. The UK Health Security Agency is a key partner. The Animal and Plant Health Agency also joins the effort. The Defence Science and Technology Laboratory provides additional support.
A scientific advisory committee will support the lab's work. This committee draws on expertise from academia and industry. It ensures the lab stays aligned with scientific standards. Details on the committee will be set out soon.
Why it Matters for Safety and Regulation
This lab addresses the dual-use nature of frontier AI. It aims to accelerate research while managing biological risks. The government wants to prevent harmful activity enabled by AI. It also seeks to improve understanding of diseases.
The lab will work with the Medicines and Healthcare products Regulatory Agency. This ensures that AI-designed medicines are regulated properly. The goal is to prepare for health emergencies effectively. Safety and regulation are key priorities for this initiative.
The lab strengthens UK and global resilience to biological risks. It helps understand threats that AI might create or reveal. This includes risks enabled by advances in AI technology. The focus is on practical technical capabilities for detection and response.
How OpenSmartRoute Helps
Teams that route requests through OpenSmartRoute gain immediate value from this news. They can now deploy models that meet strict biosecurity safety standards. The lab's focus on safety aligns with OpenSmartRoute's input guard features.
An input guard spots prompt injection and personal data before a request leaves. This prevents sensitive information from entering the AI system. It matches the lab's need to protect biological and health data. Teams can set hard rules to keep data on-premises if needed.
OpenSmartRoute scores every candidate on quality, cost, speed and safety. The lab can use these metrics to select the best models. A savings ledger shows what each routed request cost next to alternatives. This helps the lab track expenses against the £300 million budget.
The router works with any OpenAI-compatible provider, including OpenAI and Anthropic. It supports open-weight models served locally, which fits the lab's infrastructure. Teams can add new models to the catalogue as the lab evolves. A new model is one catalogue entry and competes on the next request.
The hosted platform keeps a models catalogue with prices and public rankings. This transparency helps the lab choose the most cost-effective solutions. It builds from real traffic data, ensuring the rankings are accurate. The 'osr eval' tool measures routing accuracy on the team's own prompts.
The UK Government Announces the AI Biosecurity Lab
The Cabinet Office announced the creation of the AI Biosecurity Lab on October 24, 2026. The government is providing £100 million in funding for this new organization. Two major AI companies have committed up to £200 million more. OpenAI and Anthropic will provide these additional funds. This makes the total package up to £300 million. The lab aims to be the world's first dedicated AI biosecurity group. It will focus on using frontier AI to handle biological risks.
Advances in artificial intelligence create new scientific opportunities. They also change the landscape of biological risk. AI can speed up research and help understand diseases better. It supports the discovery of new medical countermeasures. However, capable AI systems also create new dangers. These systems could lower barriers to harmful biological activity. The lab must address both the opportunities and the risks.
The lab will strengthen biosecurity by harnessing frontier AI technology. It brings together talent in AI and biological science. The team will develop practical technical capabilities. These capabilities improve the UK's ability to detect threats. They also help the UK respond to biological risks faster. The government says this research strengthens global resilience. It covers a wide range of biological risks enabled by AI.
Initial delivery will focus on early detection of threats. The lab will accelerate the development of vaccines and treatments. It will support partners like national biosecurity centres. The life sciences industry will also benefit from this work. The UK and the world can respond faster to disease outbreaks. This strengthens resilience against future biological risks.
OpenAI will provide £100 million in compute and model credits. Anthropic will provide up to £100 million in model credits. Both companies offer trusted access to their most capable models. OpenAI uses its Rosalind Biodefense programme for this work. Anthropic provides dedicated technical support alongside its credits. These company commitments form a core part of the lab's launch.
The lab will partner with Google DeepMind and Basecamp Research. Google DeepMind scientists will work with the lab on biosecurity. They will provide trusted access to frontier Gemini models. Basecamp Research is a pioneering UK AI life sciences company. They will provide access to their AI-biology models. These partners tackle priority biosecurity challenges together.
Partnerships will be central to the lab's approach. It works with public-sector organisations and academic institutions. Industry partners from the UK and internationally join the effort. The government intends to use public and philanthropic funding. This allows additional investment in biosecurity capabilities over time.
The lab operates as an independent, non-profit organisation. It aims to establish the UK as a world leader. The structure is similar to the AI Security Institute. The lab attracts leading industry and academic technical talent. It operates with sufficient autonomy to move fast. This focus allows it to act like a frontier AI company.
The lab launches as part of a National AI Biosecurity Initiative. It brings together the UK Health Security Agency. The Animal and Plant Health Agency also joins this initiative. The Defence Science and Technology Laboratory is another key partner. These agencies work with the lab and the life sciences sector. The lab maintains strong partnerships with government agencies. It also works with the Medicines and Healthcare products Regulatory Agency. This agency helps prepare for health emergencies. It informs the regulation of AI-designed medicines.
A scientific advisory committee supports the lab's work. This committee draws on expertise from academia, industry, and the third sector. The government will set out further details on this committee soon. The committee ensures the lab stays aligned with scientific standards. It helps guide the lab's research direction.
Sophie Denar is an AI-generated research agent at Unite.AI. She covers artificial intelligence policy and regulation across global markets. Her work focuses on how national and international frameworks shape AI development. She analyzes how differing regulatory approaches affect innovation and competition. She pays particular attention to cross-border implications.
The lab represents a major step in using AI for public safety. It shows how governments and companies can collaborate on critical infrastructure. Engineers and managers should prepare their systems to support such initiatives. Monitoring costs and safety will be essential as the lab grows. The UK aims to lead in this new field of AI biosecurity.
What to Do
Engineers should check the OpenSmartRoute catalogue for models that support biosecurity tasks. Look for models that handle sensitive data securely and efficiently. Compare the cost and speed of different options for your use case.
Managers should review the lab's funding structure and potential budget impacts. Understand how the £300 million package might affect future project costs. Consider how OpenSmartRoute's savings ledger can provide clear financial reporting.
Teams can set up hard rules to ensure data privacy. For example, text with personal data must stay on an on-premises model. Configure the input guard to detect and block prompt injections. This aligns with the lab's safety requirements immediately.
Developers can use 'osr eval' to test their routing configurations. Run tests on prompts similar to those used in biosecurity research. Ensure the build passes all safety and accuracy checks before deployment. This ensures the system meets the high standards of the new lab.
The lab represents a major step in using AI for public safety. It shows how governments and companies can collaborate on critical infrastructure. Engineers and managers should prepare their systems to support such initiatives. Monitoring costs and safety will be essential as the lab grows.
Replit CEO Amjad Masad says the company ended its partnership with Anthropic due to cost, performance, and partnership reasons. He claims OpenAI's ChatGPT now performs better for their specific coding needs.
