OpenAI announced the acquisition of Orchestra Bio. Orchestra Bio is a platform that helps biopharma teams coordinate the complex processes of drug development. According to a Testing Catalog report on October 21, 2026, the two sides will continue development based on OpenAI's Rosalind model and tools for life sciences research, in order to simplify these processes and bring new therapies to patients faster. There are currently no further developments confirmed by factual evidence.
According to The Information, Sam Altman told OpenAI employees that the company must decide within the next year which of three safety endgame paths to pursue. The report says the path he favors is to establish an international agency modeled on the International Atomic Energy Agency; this idea can be traced back to a proposal he put forward together with Greg Brockman and Ilya Sutskever in May 2023, under which the agency could inspect systems, require audits, and track compute and energy use. No subsequent developments have been disclosed.
According to The Information, Sam Altman told OpenAI employees that the company must decide within the next year which of three safety endgame paths to pursue. The report says the path he favors is to establish an international agency modeled on the International Atomic Energy Agency; this idea can be traced back to a proposal he put forward together with Greg Brockman and Ilya Sutskever in May 2023, under which the agency could inspect systems, require audits, and track compute and energy use.
On October 21, 2026, OpenAI research lead Mark Chen publicly welcomed Orchestra Bio to OpenAI. According to the same report, OpenAI acquired Orchestra Bio, a platform that helps biopharma teams coordinate the complex processes of drug development. Mark Chen said the two sides will use Rosalind, an AI model and toolset for life sciences research, to simplify these processes and bring new therapies to patients faster. Currently, this news has only this one source, and there are no subsequent factual developments.
On October 21, 2026, OpenAI research lead Mark Chen publicly welcomed Orchestra Bio to OpenAI. According to the same report, OpenAI acquired Orchestra Bio, a platform that helps biopharma teams coordinate the complex processes of drug development. Mark Chen said the two sides will use Rosalind, an AI model and toolset for life sciences research, to simplify these processes and bring new therapies to patients faster. Currently, this news has only this one source, and there are no subsequent factu
On October 21, 2026, OpenAI research lead Mark Chen publicly welcomed Orchestra Bio to OpenAI. According to the same report, OpenAI acquired Orchestra Bio, a platform that helps biopharma teams coordinate the complex processes of drug development. Mark Chen said the two sides will use Rosalind, an AI model and toolset for life sciences research, to simplify these processes and bring new therapies to patients faster. Currently, this news has only this one source, and there are no subsequent factual developments.
On October 20, 2026, OpenAI said its latest model has enhanced software building and migration capabilities in low-resource programming languages such as COBOL, Fortran, and Visual Basic. OpenAI said these languages underpin systems that banks, governments, and enterprises rely on, and that the progress means the model can understand decades of business logic, helping developers modernize systems without breaking the parts that already work. OpenAI also said it is building these capabilities around the challenges raised by developers and enterprise customers, and is soliciting team feedback on which languages and workflows to address next. On the same day, another report said OpenAI released a programming model, GPT-6 Astra, for legacy programming languages such as COBOL, Fortran, and Visual Basic, and said the model has been made available to developers through the OpenAI API for programming tasks in COBOL, Fortran, OCam, and Visual Basic; the report also provided score data for two evaluations, MultiLangBench and COBOL migration. The above evaluation figures come from that media report; OpenAI's official statement did not mention specific model names or scores.
On October 20, 2026, OpenAI published a post on its Alignment research blog proposing two complementary monitorability metrics in addition to g-mean2. According to the post, the pair-matched metric compares monitorability only on instances that are jointly supported by multiple models; the post-control risk metric simulates the probability that the system still produces harmful outputs after the monitor rejects an output. The post notes that g-mean2 was previously used to measure a monitor's ability to distinguish positive from negative samples, but because different models have different subsets of valid positive-sample tasks, the scores may reflect different mixtures of tasks. The post also states that if the monitor is used as a deployment control, the model can retry after being rejected, and the risk depends on the frequency of bad behavior, which outputs the monitor accepts, and the number of attempts allowed. This content currently comes from OpenAI's official blog and has not yet been cross-verified by other independent sources.
On October 20, 2026, OpenAI published a post on its Alignment research blog proposing two complementary monitorability metrics in addition to g-mean2. According to the post, the pair-matched metric compares monitorability only on instances that are jointly supported by multiple models; the post-control risk metric simulates the probability that the system still produces harmful outputs after the monitor rejects an output. The post notes that g-mean2 was previously used to measure a monitor's abi
OpenAI has launched a chat lock feature for ChatGPT. Users can set a four-digit PIN code, and on supported devices they can also use Face ID or Touch ID to protect sensitive conversations. Locked conversations are moved to a separate section, isolated from regular chat history, pinned conversations, and search results; when memory is enabled, locked conversations have their own separate memory space. The feature is being rolled out gradually to Free, Go, Plus, Pro, Business, Enterprise, and Edu users. The report also notes that the feature does not enable end-to-end encryption and does not change data retention periods. There are currently no subsequent developments.
OpenAI has launched a chat lock feature for ChatGPT. Users can set a four-digit PIN code, and on supported devices they can also use Face ID or Touch ID to protect sensitive conversations. Locked conversations are moved to a separate section, isolated from regular chat history, pinned conversations, and search results; when memory is enabled, locked conversations have their own separate memory space. The feature is being rolled out gradually to Free, Go, Plus, Pro, Business, Enterprise, and Edu
OpenAI has launched a chat lock feature for ChatGPT. Users can set a four-digit PIN code, and on supported devices they can also use Face ID or Touch ID to protect sensitive conversations. Locked conversations are moved to a separate section, isolated from regular chat history, pinned conversations, and search results; when memory is enabled, locked conversations have their own separate memory space. The feature is being rolled out gradually to Free, Go, Plus, Pro, Business, Enterprise, and Edu users. The report also notes that the feature does not enable end-to-end encryption and does not change data retention periods. There are currently no subsequent developments.
OpenAI announced the acquisition of Orchestra Bio. Orchestra Bio is a platform that helps biopharma teams coordinate the complex processes of drug development. According to a Testing Catalog report on October 21, 2026, the two sides will continue development based on OpenAI's Rosalind model and tools for life sciences research, in order to simplify these processes and bring new therapies to patients faster. There are currently no further developments confirmed by factual evidence.
According to The Information, Sam Altman told OpenAI employees that the company must decide within the next year which of three safety endgame paths to pursue. The report says the path he favors is to establish an international agency modeled on the International Atomic Energy Agency; this idea can be traced back to a proposal he put forward together with Greg Brockman and Ilya Sutskever in May 2023, under which the agency could inspect systems, require audits, and track compute and energy use. No subsequent developments have been disclosed.
OpenAI is accelerating its push for about $30 billion in financing. According to Reuters, OpenAI is aiming to complete this funding round at a valuation of more than $1.4 trillion before Anthropic goes public, a valuation at least 64% higher than the $852 billion it reached in March. OpenAI hopes to lock in commitments before Anthropic files its S-1 and lists in mid-November, so as to avoid a direct comparison. Subsequent reports said existing investors such as SoftBank and Abu Dhabi's MGX are in talks to participate; OpenAI told investors that growth accelerated after the release of GPT-6 Astra, with annualized revenue already reaching $50 billion and expected to rise to $70 billion by the end of the year. The financing is still in progress.
On October 20, 2026, OpenAI said its latest model has enhanced software building and migration capabilities in low-resource programming languages such as COBOL, Fortran, and Visual Basic. OpenAI said these languages underpin systems that banks, governments, and enterprises rely on, and that the progress means the model can understand decades of business logic, helping developers modernize systems without breaking the parts that already work. OpenAI also said it is building these capabilities around the challenges raised by developers and enterprise customers, and is soliciting team feedback on which languages and workflows to address next. On the same day, another report said OpenAI released a programming model, GPT-6 Astra, for legacy programming languages such as COBOL, Fortran, and Visual Basic, and said the model has been made available to developers through the OpenAI API for programming tasks in COBOL, Fortran, OCam, and Visual Basic; the report also provided score data for two evaluations, MultiLangBench and COBOL migration. The above evaluation figures come from that media report; OpenAI's official statement did not mention specific model names or scores.
On October 20, 2026, OpenAI said its latest model has enhanced software building and migration capabilities in low-resource programming languages such as COBOL, Fortran, and Visual Basic. OpenAI said these languages underpin systems that banks, governments, and enterprises rely on, and that the progress means the model can understand decades of business logic, helping developers modernize systems without breaking the parts that already work. OpenAI also said it is building these capabilities aro
On October 20, 2026, OpenAI published a post on its Alignment research blog proposing two complementary monitorability metrics in addition to g-mean2. According to the post, the pair-matched metric compares monitorability only on instances that are jointly supported by multiple models; the post-control risk metric simulates the probability that the system still produces harmful outputs after the monitor rejects an output. The post notes that g-mean2 was previously used to measure a monitor's ability to distinguish positive from negative samples, but because different models have different subsets of valid positive-sample tasks, the scores may reflect different mixtures of tasks. The post also states that if the monitor is used as a deployment control, the model can retry after being rejected, and the risk depends on the frequency of bad behavior, which outputs the monitor accepts, and the number of attempts allowed. This content currently comes from OpenAI's official blog and has not yet been cross-verified by other independent sources.
On October 21, 2026, OpenAI released a major update to ChatGPT's Chrome extension. According to TechSpot, the update adds two modes, Chat and Work, which can read the content of open tabs, control the browser to send messages, download files, group tabs by topic, and fill out forms using online information or local files, with the user manually reviewing before submission. Users can also directly execute browser tasks by using @Chrome in the ChatGPT app or website, and can limit access to specific tabs; when accessing chat history or new websites, authorization will be requested, offering four options: once only, this site only, all sites, and deny. The new features are available for free in Chromium browsers such as Chrome, Edge, Brave, Opera, and Vivaldi, but are not yet supported in Firefox and Safari, and the extension can be downloaded from the Chrome Web Store. There are currently no subsequent developments.
On October 21, 2026, OpenAI released a major update to ChatGPT's Chrome extension. According to TechSpot, the update adds two modes, Chat and Work, which can read the content of open tabs, control the browser to send messages, download files, group tabs by topic, and fill out forms using online information or local files, with the user manually reviewing before submission. Users can also directly execute browser tasks by using @Chrome in the ChatGPT app or website, and can limit access to specif