OpenAI plans to launch Orion, its next frontier model, by December, The Verge has learned.

Unlike the release of OpenAI’s last two models, GPT-4o and o1, Orion won’t initially be released widely through ChatGPT. Instead, OpenAI is planning to grant access first to companies it works closely with in order for them to build their own products and features, according to a source familiar with the plan.

Another source tells The Verge that engineers inside Microsoft — OpenAI’s main partner for deploying AI models — are preparing to host Orion on Azure as early as November. While Orion is seen inside OpenAI as the successor to GPT-4, it’s unclear if the company will call it GPT-5 externally. As always, the release plan is subject to change and could slip. OpenAI and Microsoft declined to comment for this story.

Orion has been teased by an OpenAI executive as potentially up to 100 times more powerful than GPT-4; it’s separate from the o1 reasoning model OpenAI released in September. The company’s goal is to combine its LLMs over time to create an even more capable model that could eventually be called artificial general intelligence, or AGI.

It was previously reported that OpenAI was using o1, code named Strawberry, to provide synthetic data to train Orion. In September, OpenAI researchers threw a happy hour to celebrate finishing training the new model, a source familiar with the matter tells The Verge.

That timing lines up with a cryptic post on X by OpenAI CEO Sam Altman, in which he said he was “excited for the winter constellations to rise soon.” If you ask ChatGPT o1-preview what Altman’s post is hiding, it will tell you that he’s hinting at the word Orion, which is the winter constellation that’s most visible in the night sky from November to February (but it also hallucinates that you can rearrange the letters to spell “ORION”).

The release of this next model comes at a crucial time for OpenAI, which just closed a historic $6.6 billion funding round that requires the company to restructure itself as a for-profit entity. The company is also experiencing significant staff turnover: CTO Mira Murati just announced her departure along with Bob McGrew, the company’s chief research officer, and Barret Zoph, VP of post training.

If Batman: Arkham Shadow is tempting you to finally pick up a Meta Quest 3 or Meta Quest 3S, or you already own one of Meta’s best VR headsets, then this might sweeten the deal. Meta has revealed the dates of the 52 free NBA and WNBA matches it’s hosting in Horizon Worlds for the 2024-2025 season.

To catch you up to speed, Meta allows its headset owners (in supported regions) to jump into the NBA Arena in Horizon Worlds to virtually sit courtside during live matches. If you miss the game, you can also tune into immersive highlights for 24 hours after the game has ended. Additionally, you can play mini-games and hang out with friends like you would in other Horizon Worlds.

“While both public and private sectors face cyber security challenges, the public sector tends to be more exposed to the threats in an ever-evolving landscape,” consultancy major KPMG said in a recent report. A Palo Alto study last year said that 67 per cent of Indian government and critical public sector installations had a 50 per cent increase in cyber attacks

It is no exaggeration. From the country’s premier medical institution—the All India Institute of Medical Sciences (AIIMS) came under not one, but two malware attacks over the course of one year—to premier banks losing money, India, and particularly its public sector, remain vulnerable to cyber crimes in what is one of the top five cyber attacked countries in the world.

ALSO READ | India’s defence, other govt departments fell prey to cyber attacks in 2024, Telegram a hotbed: Report

“India’s public sector is the Achilles’ heel of our national cyber security,” said Trishneet Arora, founder and CEO of the cyber security company TAC Securities. “The absence of an actionable risk management system and outdated infrastructure leave critical systems vulnerable to cyber threats.”

Banks, including public sector banks, are a particularly high-value target. According to data, there were 248 successful breaches of Indian banks by cyber attackers in the four year period between 2018 and 2022, primarily card information leakage and theft. However, in a serious incident in November last year, UCO Bank had reported erroneous crediting of more than 800 crore rupees via IMPS. The bank later managed to recover nearly 80 per cent of the amount through actions like freezing accounts.

After a cyber security and information technology examination, or CSITE, identified vulnerabilities in certain Indian banks, the Reserve Bank of India (RBI) had in March this year cautioned banks to adopt cyber security measures. RBI’s deputy governor T. Rabi Sankar, speaking at a banking conference in February, had called on banks to upgrade their encryption systems to counter artificial intelligence (AI)-spawned attacks.

RBI has also instituted a dedicated Cyber Security Framework for Scheduled Commercial Banks.

For India’s public sector undertakings (PSU), the problems stem from “legacy systems, staff training, bureaucratic complexities, relying on third parties, difficulties in continuous monitoring and real-time threat detection (and) cultural resistance to change, scarcity of specialised cyber security personnel…(all) leaving these institutions vulnerable to evolving cyber risks,” according to Ruchin Kumar, vice president (South Asia) of Futurex, a US-based cyber security company.

So what can be done? Modernising IT infra and complying to cyber security regulations would be a no-brainer, but many PSUs still don’t allocate enough budget, or specialised personnel, for this. “Furthermore, strengthening third-party and supply chain security, promoting inter-agency collab and enhancing threat detection capabilities are vital components of this approach,” added Ruchin Kumar. With cyber threats always evolving, it is imperative that PSUs invest in continuous monitoring tools and stay updated. They also need to enhance security by following government updates, including encryption and tokenisation to protect sensitive data, as well as deploying hardware security modules (HSMs) to manage and protect cryptographic keys.

ALSO READ | ‘Data breach, ransomware threats amount to cyberattack on India, not trolling Prime Minister or LoP online’

The situation is even more critical for public sector banks, as they deal in millions of financial transactions daily. “Cyber criminals obtain data from social networks and are also learning new technologies to make cyber attacks,” pointed out Rajendra K. Sinha, professor and chairperson, Centre of Excellence in Banking, JAGSoM. “Further, they use methods that are not easily detected by endpoint protection code.”

So what can the hapless ordinary citizen who is a bank customer do? Sinha has some tips in addition to training and awareness of bank staff and customers. “Precautionary measures include changing passwords regularly with strong password, removing personal information from social media, and not opening emails from an unknown source.”

Strands is a brand new daily puzzle from the New York Times. A trickier take on the classic word search, you’ll need a keen eye to solve this puzzle.

Like Wordle, Connections, and the Mini Crossword, Strands can be a bit difficult to solve some days. There’s no shame in needing a little help from time to time. If you’re stuck and need to know the answers to today’s Strands puzzle, check out the solved puzzle below.

How to play Strands

Hint for today’s Strands puzzle

Today’s Strand answers

Today’s spanagram

We’ll start by giving you the spangram, which might help you figure out the theme and solve the rest of the puzzle on your own:

Today’s Strands answers

• BARK
• MEOW
• GRUNT
• SQUAWK
• GROWL
• PURR
• CHIRP
• SQUEAK

Google Photos is undoubtedly one of the best photo management apps, thanks to new and enhanced features. That said, it can’t clearly tell if an image has AI edits. Earlier this month, some reports hinted that Google Photos would soon help identify AI-generated images. Now, the company announced that Google Photos will soon tell you if an image has been through AI edits.

Google Photos’ “AI Info” section will clearly label images with AI edits

Starting next week, Google Photos will clearly label images that went through edits using AI tools like Magic Editor, Magic Eraser, and Zoom Enhance. Google, in the announcement post, says that it is making this change to increase transparency and help users understand more about AI-edited images. With the growing popularity of Gen AI tools, this move was much needed.

While the metadata of an image file used to include the use of AI, it wasn’t easily accessible to users. However, that’ll change starting next week. Google Photos will include a new “AI Info” section alongside image details like file name, backup status, and location. This will allow users to easily know if AI has altered the images.

The new section will be available on Google Photos web and mobile app

It’s worth noting that Google Photos will label the edits involving generative AI. That’s not all, Google also says “We will also use IPTC metadata to indicate when an image is composed of elements from different photos using non-generative features.” Here, Google is referring to “Best Take” and “Add Me” features.

The “AI Info” section will be visible in the image details view of the Google Photos web and mobile app. While this is a welcome change, Google has yet to come up with something that will let users immediately identify AI-edited images.

The White House issued its first national security memorandum outlining the use of artificial intelligence for the military and intelligence agencies. The White House also shared a shortened copy of the memo with the public.

The new memo sets up guidelines for military and intelligence agencies for using AI in its day-to-day operations. The memo sets a series of deadlines for agencies to study the applications and regulations of AI tools, most of which will lapse following President Biden’s term. The memo also aims to limit “the most dystopian possibilities, including the development of autonomous weapons,” according to the .

National Security Adviser Jake Sullivan announced the new directive today at as part of a talk on AI’s presence in government operations. Sullivan has been one of the President’s most vocal proponents for examining the benefits and risks of AI technology. He also raised concerns about China’s use of AI to control its population and spread misinformation and how the memo can spark conversations with other countries grappling with implementing its own AI strategies.

The memorandum establishes some hard edges for AI usage especially when it comes to weapons systems. The memo states that AI can never be used as a decision maker for launching nuclear weapons or assigning asylum status to immigrants coming to the US. It also prohibits AI from tracking anyone based on their race or religion or determining if a suspect is a known terrorist without human intervention.

The memo also lays out protections for private-sector AI advance as “national assets that need to be protected…from spying or theft by foreign adversaries,” according to the Times. The memorandum orders intelligence agencies to help private companies working on AI models secure their work and provide updated intelligence reports to project their AI assets.