Apple today announced its largest-ever spend commitment, with plans to spend and invest more than $500 billion in the U.S. over the next four years.
This new pledge builds on Apple’s long history of investing in American innovation and advanced high-skilled manufacturing, and will support a wide range of initiatives that focus on artificial intelligence, silicon engineering, and skills development for students and workers across the country.
“We are bullish on the future of American innovation, and we’re proud to build on our long-standing U.S. investments with this $500 billion commitment to our country’s future,” said Tim Cook, Apple’s CEO.
“From doubling our Advanced Manufacturing Fund, to building advanced technology in Texas, we’re thrilled to expand our support for American manufacturing. And we’ll keep working with people and companies across this country to help write an extraordinary new chapter in the history of American innovation.”
Apple suppliers already manufacture silicon in 24 factories across 12 states. Pictured: Texas Instruments’ new semiconductor wafer fabrication plant in Lehi, Utah. Photo: Apple
As part of this package of U.S. investments, Apple and partners will open a new advanced manufacturing facility in Houston to produce servers that support Apple Intelligence, the personal intelligence system that helps users write, express themselves, and get things done.
Apple will also double its U.S. Advanced Manufacturing Fund, create an academy in Michigan to train the next generation of U.S. manufacturers, and grow its research and development investments in the U.S. to support cutting-edge fields like silicon engineering.
The $500 billion commitment includes Apple’s work with thousands of suppliers across all 50 states, direct employment, Apple Intelligence infrastructure and data centers, corporate facilities, and Apple TV+ productions in 20 states. Apple remains one of the largest U.S. taxpayers, having paid more than $75 billion in U.S. taxes over the past five years, including $19 billion in 2024 alone.
Today, Apple supports more than 2.9 million jobs across the country through direct employment, work with U.S.-based suppliers and manufacturers, and developer jobs in the thriving iOS app economy.
Apple has announced a new iPhone which brings artificial intelligence (AI) features at a lower cost than its flagship handsets.
The iPhone 16e has the same processor as the larger iPhone 16, Apple said, with similar storage options, though a lower spec elsewhere, including fewer cameras.
Apple has been struggling to find a new product that excites consumers – sales of iPhones dropped at the end of last year.
It will be hoping that bringing enhanced AI functionality to a less expensive phone will address that – however analysts have been cautious about the sales boost such tools bring.
Its name is clearly a nod to its iPhone SE series, which were released from 2016 to 2022, and similarly priced significantly lower.
Apple said the iPhone 16e will be available for pre-order from February 21 in 59 countries.
It will launch in the UK for £599, which is £200 less than the iPhone 16 – but more than double what the original iPhone SE went for when it launched in 2016.
“This now becomes one of the most affordable powerful iPhones now on the market,” industry analyst Paolo Pescatore told BBC News.
“The move should help accelerate adoption and especially its foray into AI with Apple Intelligence.
“Apple’s trust and credibility is critical – this alone will help drive sales and lure users from rival devices and platforms.”
Apple Intelligence
Much of the conversation around the new handset will likely centre around its power, with Apple electing to use the same A18 chip behind its more expensive devices.
This means the 16e will be capable of playing the same games and running the same apps as other iPhones – though AI is likely at the heart of this decision.
Apple boss Tim Cook said in the announcement the new model featured “the performance, intelligence and privacy” Apple fans “expect” from the firm.
And he said the Apple Intelligence features on the device would “help you save time, quickly get more things done, and express yourself in new ways”.
The firm introduced its spin on the tech – Apple Intelligence – with this series of devices, which includes new tools for writing and incorporating OpenAI’s chatbot ChatGPT into Siri.
It hasn’t always gone well, with the firm at one point suspending its AI-generated news alerts after they created false headlines attributed to news organisations including the BBC.
It now presents the summaries in italics.
Apple said its new phone is “built for Apple Intelligence”, and pointed to certain features of the tech, like an easy way to clean up photos or search your image library.
Other phone manufacturers have similar features on their devices – though the iPhone 16e will be by far the cheapest way to access AI on an Apple handset.
“The iPhone 16e generates a new revenue stream for Apple, and this will be particularly noticeable in key markets like India, where iPhones are out of reach for most people,” said Forrester principal analyst Dipanjan Chatterjee.
“There is also a second-order effect of cheaper devices like the iPhone 16e, bringing new customers into the Apple ecosystem.
“We’ve seen a limited appetite among many of the installed base to upgrade from previous versions, but the new phone reduces the cost hurdle of joining the Apple Intelligence bandwagon.”
Apple’s current offerings are also underwhelming in light of the company’s prolific investments in research and development. Cory Johnson, Epistrophy Capital Research chief market strategist, said Apple has spent $189bn on R&D over the last decade alone.
“All we have to show for that is the HomePod and $3,500 ski goggles,” Johnson said. “AI should be right in Apple’s wheelhouse. But Apple fanboys, fangirls, and investors are right to be disappointed so far.”
WhatsApp has introduced a new feature that allows users to customise their chat theme colour, background and bubbles.
The feature, which is available for both iOS and Android users allows them to make chats unique with colourful chats and new wallpapers.
“You asked for it. Here it is! We are introducing chat themes, so you can make your chats uniquely yours with colourful chat bubbles and new wallpapers,” they said in a blog.
1. Changing the wallpapers
The rolled-out pre-set themes have been there, but now, WhatsApp says users have 30 options to choose from, set a plain colour or upload a new background for the chat from the photo gallery.
WhatsApp said that users can choose to apply a theme to all of their chats or choose a theme for a specific conversation.
These themes are only visible to the user and can be applied to the WhatsApp Channels as well.
The themes also allow users to customise the background brightness and even change the colour of outgoing messages.
However, the colour of incoming messages will either be white or grey, depending on whether you are using a light or dark theme in the app.
2. Changing the theme
To change the theme of all your chats and channels:
1. Go to Settings then Chats.
2. Click on the default chat theme.
3. To customise each chat individually, click the chat name at the top of the chat screen (on iOS), or click the 3-dot menu (on Android).
4. Click the chat theme and customise it accordingly.
WhatsApp said the features are now starting to roll out and will be available globally over the coming weeks.
Elon Musk said on Thursday his AI chatbot, and ChatGPT challenger, Grok 3, is in the final stages of development and will be released in about a week or two.
“Grok 3 has very powerful reasoning capabilities, so in the tests that we’ve done thus far, Grok 3 is outperforming anything that’s been released, that we’re aware of, so that’s a good sign,” he said in a video call addressing the World Governments Summit in Dubai.
The billionaire tech mogul founded xAI as a challenger to Microsoft-backed OpenAI and Alphabet’s Google. Musk also co-founded OpenAI.
On Monday, a consortium of investors led by Musk said it had offered $97.4 billion to buy the assets of OpenAI’s nonprofit, in another salvo from the world’s richest man against the artificial intelligence startup.
OpenAI has said it wants to become a for-profit organization to secure the capital needed for developing the best AI models.
Musk sued OpenAI CEO Sam Altman and others in August and has asked a U.S. district judge to block OpenAI’s attempt to transition to a for-profit entity. OpenAI said this week Musk’s bid clashes with his lawsuit.
“I think the evidence is there in that OpenAI has gotten this far while having at least a sort of dual profit, non-profit role. What they’re trying to do now is to completely delete the non-profit, and that seems really going too far.”
Musk, who was appointed by U.S. President Donald Trump to oversee the so-called Department of Government Efficiency aimed at dramatically reducing the size of the federal workforce, said government spending could be reduced by $1 trillion or more.
“Maybe the economy could grow at 4 or 5% potentially, in terms of real useful goods and services output, and government spending can be reduced by about 3 or 4% of the economy, about maybe a trillion dollars or more, and the net effect of that would be no inflation from 2025 to 2026 so that would be quite remarkable,” Musk said.
UAE AI Minister Omar Al Olama, who was interviewing Musk at the conference, said they would partner on “Dubai Loop”, an underground high-speed transport system that Musk likened to a wormhole. Al Olama did not give details.
Turning to international affairs, Musk told the Middle East audience the United States has been “pushy” in the past and it should “mind its own business”.
“I think we should, in general leave other countries to their own business,” he said.
Trump has enraged the Arab world by saying the U.S. would take over the Gaza strip, resettle its Palestinian inhabitants and turn it into the “Riviera of the Middle East”.
A consortium led by Elon Musk said on Monday it has offered $97.4 billion (Sh1.3 trillion) to buy the nonprofit that controls OpenAI, another salvo in the billionaire’s fight to block the artificial intelligence startup from transitioning to a for-profit firm.
Musk’s bid is likely to ratchet up longstanding tensions with OpenAI CEO Sam Altman over the future of the ChatGPT maker at the heart of a boom in generative AI technology. Altman on Monday promptly posted on X: “no thank you but we will buy twitter for $9.74 billion (Sh1.3 trillion) if you want.”
Musk cofounded OpenAI with Altman in 2015 as a nonprofit, but left before the company took off. He founded the competing AI startup xAI in 2023.
Musk, the CEO of Tesla and owner of tech and social media company X, is a close ally of President Donald Trump. He spent more than a quarter of a billion dollars to help elect Trump, and leads the Department of Government Efficiency, a new arm of the White House tasked with radically shrinking the federal bureaucracy. Musk recently criticized a $500 billion (Sh 64.3 trillion) OpenAI-led project announced by Trump at the White House.
OpenAI is now trying to transition into a for-profit from a nonprofit entity, which it says is required to secure the capital needed for developing the best AI models.
Musk sued Altman and others in August last year, claiming they violated contract provisions by putting profit ahead of the public good in the push to advance AI. In November, he asked a US district judge for a preliminary injunction blocking OpenAI from converting to a for-profit structure.
Musk’s lawsuit against OpenAI and Altman says the founders originally approached him to fund a nonprofit focused on developing AI to benefit humanity, but that it was now focused on making money.
“It’s time for OpenAI to return to the open-source, safety-focused force for good it once was,” Musk said in a statement on Monday. “We will make sure that happens.”
Altman told staff in a message that the company’s board of directors intends to make clear it has no interest in Musk’s “supposed bid“, according to a report by The Information on Monday.
Musk and OpenAI backer Microsoft did not immediately respond to requests for comment.
The consortium led by Musk includes his AI startup xAI, Baron Capital Group, Emanuel Capital and others.
xAI could merge with OpenAI following a deal, according to the Wall Street Journal, which first reported Musk’s offer earlier on Monday. xAI recently raised $6 billion from investors at a valuation of $40 billion, sources have told Reuters.
‘Throwing a wrench’
“This (bid) is definitely throwing a wrench in things,” said Jonathan Macey, a Yale Law School professor specializing in corporate governance.
“The nonprofit is supposed to take money to do whatever good deeds, and if OpenAI prefers to sell it to somebody else for less money, it’s a concern for protecting the interests of the beneficiaries of the not-for-profit.”
OpenAI was valued at $157 billion (Sh20.2 trillion) in its last funding round, cementing its status as one of the most valuable private companies in the world. SoftBank Group is in talks to lead a funding round of up to $40 billion in OpenAI at a valuation of $300 billion (Sh38.6 trillion), including the new funds, Reuters reported in January.
Aside from any antitrust implications, a deal this size would need Musk and his consortium to raise enormous funds.
Musk’s stock in Tesla is valued at roughly $165 billion (Sh21.2 trillion), according to LSEG data, but his leverage with banks is likely to be thin after his $44 billion buyout of X, which was then called Twitter, in 2022.
To finance such a bid, Musk could sell part of his stake in Tesla or take a loan against his stake, or use his stake in rocket company SpaceX that is worth tens of billions of dollars as collateral, according to an uninvolved investment banker, who requested anonymity.
“Musk’s offer to buy OpenAI’s nonprofit should significantly complicate OpenAI’s current fundraising and the process of converting into a for-profit corporation,” said Gil Luria, analyst at D.A. Davidson.
“The offer seems to be backed by more credible investors … OpenAI may not be able to ignore it. It will be the fiduciary responsibility of OpenAI’s board to decide whether this is a better offer, which could call into question the offer from SoftBank.”
Billionaire Elon Musk said that he was not interested in purchasing TikTok, the popular short-video app that the United States has been trying to ban over national security concerns with its Chinese owner ByteDance.
Musk’s comments, opens new tab, made in late January, were released online Saturday by The WELT Group, a part of the German media company Axel Springer SE, which hosted a summit, where the Tesla, opens new tab chief joined the conference via video.
“I have not put in a bid for TikTok,” Musk said a week after U.S. President Donald Trump said he was open to Musk buying the ByteDance-owned app if he wanted to do so.
“I don’t have any plans for what I would do if I had TikTok,” Musk said, adding that he does not use the short video app personally, and was not familiar with the app’s format.
“I’m not chomping at the bit to acquire TikTok, I do not acquire companies in general, it’s quite rare,” Musk said, adding that his billion-dollar acquisition of social media platform Twitter, now called X, was unusual.
“I usually build companies from scratch,” Musk said.
The Republican president signed an executive order seeking to delay the enforcement of a ban on the popular short-video app that was slated to be shuttered on January 19.
ByteDance was given the January deadline to sell the U.S. assets of TikTok or face a U.S. ban, following lawmakers’ concerns that the app poses national security risks because China could compel the company to share the data of its U.S. users. TikTok has denied that it has or ever would share U.S. user data.
Apple and Google have not reinstated TikTok to their app stores since a U.S. law took effect. TikTok said on Friday that it was allowing U.S. Android users to download and connect to the app through package kits on its website, in an effort to circumvent restrictions on the popular platform in the country.
Trump has said that he was in talks with multiple people over TikTok’s purchase and would likely have a decision on the app’s future this month. It has about 170 million American users.
This week, the president signed an executive order to create a sovereign wealth fund within the year, saying it could potentially buy TikTok.
ByteDance has previously denied any plans to sell TikTok.
Trump saving TikTok represents a reversal in stance from his first term in office when he unsuccessfully sought to ban the app over concerns the company was sharing Americans’ personal information with the Chinese government.
More recently, Trump has said he has “a warm spot in my heart for TikTok,” crediting the app with helping him win over young voters in the 2024 presidential election.
ByteDance and TikTok did not immediately respond to a request for comment outside regular business hours.
The official X (formerly Twitter) account of the Directorate of Criminal Investigations (DCI) Kenya has fallen victim to hackers, who appear to be using the platform to push a cryptocurrency scam.
The account, which typically disseminates security and law enforcement updates, has been posting content promoting a dubious crypto scheme, alerting followers to potential fraud.
The breach was noticed when the DCI Kenya account deviated from its usual posts, instead sharing details about a cryptocurrency investment opportunity.
One of the messages posted on the account urged local media to support the launch of a blockchain project, while another promoted the distribution of tokens in exchange for likes and retweets.
DCI confirms the hack
However, the DCI seems to have regained control, as the posts have since been deleted.
The DCI has since issued the following statement regarding the hack on their X account:
“For some moment this evening, we experienced a cyber-attack on the DCI digital platforms (X and Facebook), but have since regained full control. During the short period, the cyber criminals who attempted to take over the accounts posted the information captured on the screenshot below.” The statement reads.
During the brief period when the accounts were compromised, the hackers managed to post misleading information, which was captured in a screenshot and referenced in the DCI’s statement. The DCI explicitly clarified that this information was “FAKE” and did not originate from them.
“The information is therefore FAKE and not from the DCI. A scrupulous interrogation into the criminal activity has been activated to bring to book the perpetrators.” It added.
This marks another high-profile case of cybercrime in Kenya, following closely on the heels of a similar attack on the Kenya Broadcasting Corporation (KBC) account just last week.
In the KBC incident, hackers not only took over the account but also swiftly changed the handle to “DeepseekOnSoI”, named after the AI chatbot DeepSeek, which mimics the functionality of well-known AI like ChatGPT. This rebranding was evidently an attempt to leverage the chatbot’s recognition for misleading potential investors.
This incident is part of a growing global trend where high-profile social media accounts, including those on YouTube, X, and other platforms, are hijacked for similar fraudulent activities. Cybersecurity experts have long warned about the increasing sophistication of such scams, where attackers often rebrand accounts to impersonate well-known entities or celebrities to gain trust and credibility.
Scammers would also hold the accounts until they’re paid. Ransomware is common especially with big corporates that are targeted by hackers and part with millions.
How Crypto scams works
According to cybersecurity trends, the attack method often involves phishing emails or malware designed to steal session cookies. This allows hackers to bypass even two-factor authentication measures. Once they gain control, the account is used to broadcast live streams or posts promising high returns on cryptocurrency investments. However, the funds ultimately disappear into the hands of the scammers.
U.S. tech giant OpenAI on Monday unveiled a ChatGPT tool called “deep research” ahead of high-level meetings in Tokyo, as China’s DeepSeek chatbot heats up competition in the AI field.
Artificial intelligence newcomer DeepSeek has sent Silicon Valley into a frenzy, with its high performance and supposed low cost prompting calls for US developers to go faster.
OpenAI, whose ChatGPT fronted generative AI’s emergence into public consciousness in 2022, said its new tool “accomplishes in tens of minutes what would take a human many hours”.
“Deep research is OpenAI’s next agent that can do work for you independently — you give it a prompt, and ChatGPT will find, analyze, and synthesize hundreds of online sources to create a comprehensive report at the level of a research analyst,” it said in a statement.
In a live streamed video announcement, OpenAI researchers showed how the tool can synthesize web search data to help recommend ski equipment to buy for a snow holiday in Japan.
OpenAI chief Sam Altman is in Tokyo to meet Japan’s Prime Minister Shigeru Ishiba later Monday along with Masayoshi Son, head of Japanese tech investment behemoth SoftBank Group.
SoftBank and OpenAI are part of the Stargate drive announced by US President Donald Trump to invest up to $500 billion in artificial intelligence infrastructure in the United States.
Ishiba is expected to visit Washington to meet Trump for the leaders’ first in-person meeting later this week.
– ‘New kind of hardware‘ –
On Monday afternoon, Altman and Son will hold a forum in Tokyo with around 500 businesses at which they are expected to announce plans to boost Japan’s AI infrastructure.
The Nikkei business daily reported that this will include building AI data centres and power plants to run them, without specifying the scale of the investment required.
Separately, Altman told the Nikkei he wants to develop “a new kind of hardware” using artificial intelligence in partnership with Apple’s former chief design officer Jony Ive.
But Altman indicated that it would take several years to unveil a prototype, the Nikkei said.
Altman also told the newspaper that DeepSeek is “a good model” that highlights the serious competition for AI reasoning technology, but that its “capability level isn’t new”.
DeepSeek’s performance has sparked a wave of accusations that it has reverse-engineered the capabilities of leading US technology, such as the AI powering ChatGPT.
Last week OpenAI warned that Chinese companies are actively attempting to replicate its advanced AI models, prompting closer cooperation with US authorities.
The Business Registration Services (BRS) has suffered a significant data breach following a cyber-attack, potentially exposing sensitive information about private companies to the public.
The breach, which occurred on the night of Friday, January 31, has raised serious concerns about the security of confidential data held by government agencies.
A source close to the matter confirmed the breach, revealing that BRS executives were locked in crisis meetings for most of Saturday, February 1, to address the fallout. The source, who spoke on condition of anonymity due to restrictions on speaking to the media, suggested that the breach may have involved an internal actor.
“We still can’t say who is behind the breach, but it looks like the intent is sabotage because the nature of the breach suggests an internal actor,” the source said.
Data Exposed, Dark Web Links Confirmed
The full extent of the stolen data remains unclear, but there are confirmed reports that the compromised information is being sold on the dark web, a hidden part of the internet often used for illegal activities.
Kenya Insights has verified that the leaked data, hosted on a dark web site, includes records of all registered Kenyan companies dating back to 1967. The dump contains confidential information such as the names and contact details of company owners, directors, and beneficial owners.
Data-Rich Target
The BRS is one of the most data-rich entities within the Kenyan government, holding critical information on all registered companies, including their owners, beneficial owners, and directors. This data is typically accessible only through a paid service, but the breach has potentially made it available to anyone, bypassing the usual safeguards.
The agency’s online database, which allows the public to access such information, is currently down and inaccessible. This has raised suspicions that the attackers may have deliberately taken the system offline as part of their operation.
Additionally, the Office of the Official Receiver, which operates under the BRS, maintains records of companies in financial distress. It is feared that this sensitive data may also have been compromised in the breach.
Motive Remains Unclear
While the motive behind the attack is still unknown, sources indicate that authorities have ruled out ransomware as a likely cause. Ransomware attacks typically involve hackers demanding payment in exchange for restoring access to stolen data. In this case, the breach appears to have been aimed at exposing sensitive information rather than financial extortion.
Legal and Regulatory Implications
Under Kenya’s data protection laws, organizations are required to assess the extent of any data breach, notify affected parties, and take steps to contain the situation. The BRS is expected to issue a formal statement once the full scope of the breach is understood.
This incident marks the first major data breach involving a government entity in over a year, following a cyberattack on Kenya Airways in late 2023, which resulted in the loss of significant customer data.
UPDATE: BRS Confirms Data Breach, Investigation Underway
The BRS has officially confirmed reports of a data breach affecting its company registry database.
In a statement on Sunday, February 2, BRS Director General Kenneth Gathuma acknowledged the breach and stated that the agency had initiated an immediate response.
“Upon receiving this information, we immediately activated our Incident Response Plan, launched a comprehensive investigation, and notified the relevant regulatory authorities,” the statement read.
BRS further stated that its cybersecurity experts are working closely with law enforcement, investigative agencies, and cybersecurity partners to determine the extent of the breach and implement necessary containment measures.
“Our cybersecurity experts, in collaboration with our cybersecurity partner, law enforcement, and investigative agencies, are assessing the scope of the incident, determining any potential impact, and implementing necessary containment and mitigation measures,” the agency added.
BRS Director General Kenneth Gathuma.
However, BRS noted that it is still in the process of verifying the details of the breach, including the nature and impact of any compromised data.
“At this stage, we are still verifying the details of the alleged breach, including the nature and extent of any compromised data,” the statement continued.
The agency assured stakeholders that affected parties would be directly engaged once the investigation is concluded.
Strengthening Security Measures
To mitigate further risk, BRS has implemented additional security measures to reinforce its cybersecurity infrastructure and prevent future incidents.
Additionally, the agency has pledged to maintain transparency throughout the investigation process, promising regular updates to the public and business stakeholders.
“Once the investigation is complete, we will provide an update and directly engage with any affected parties,” BRS stated.
BRS has also called for cooperation from all relevant parties as it works toward a swift resolution.
“We want to assure all stakeholders that the security and integrity of the company registry remain our top priority. As a precautionary measure, we have strengthened our security protocols to safeguard our systems and prevent future incidents,” added Director General Gathuma.
The investigation is ongoing, with further updates expected as new information becomes available.
There is a good chance that inside your mobile phone is a miniscule amount of a metal that started its journey buried in the earth of eastern Democratic Republic of Congo, where a war is currently raging.
It may even be directly connected to the M23 rebel group that made global headlines this week.
The tantalum within your device weighs less than half of the average garden pea but is essential for the efficient functioning of a smartphone, and almost all other sophisticated electronic devices.
The unique properties of this rare, blue-grey, lustrous metal – including being able to hold a high charge compared to its size, while operating in a range of temperatures – make it an ideal material for tiny capacitors, which temporarily store energy.
It is also mined in Rwanda, Brazil and Nigeria but at least 40% – and maybe more – of the element’s global supply comes from DR Congo and some of the key mining areas are now under the control of the M23.
The current wave of fighting has been going on for months, but the rebels grabbed attention with Sunday’s assault on the vital trading and transport hub of Goma. The city, bordering Rwanda, is a regional centre for the mining business
Over the past year, the M23 has made rapid advances across the mineral-rich east of DR Congo, taking areas where coltan – the ore from which tantalum is extracted – is mined.
Like scores of other armed groups operating in the area, the M23 began as an outfit defending the rights of an ethnic group perceived to be under threat. But as its territory has expanded, mining has become a crucial source of income, paying for fighters and weapons.
Last April, it seized Rubaya, the town at the heart of the country’s coltan industry.
Mineral extraction in this region is not in the hands of multinational conglomerates – instead thousands of individuals toil in open pits that honeycomb the landscape, or underground, in extremely unsafe and unhealthy conditions.
This aerial shot from Rubaya taken in 2014 shows how the coltan operation worked at one mine
They are part of a complex, and yet informal, network that sees the rocks removed from the ground using shovels, brought to the surface, crushed, washed, taxed, sold and then exported to be further purified and eventually smelted.
Once the M23 moved into Rubaya, the rebels established what a UN group of experts described as a “state-like administration”, issuing permits to the diggers and traders and demanding an annual fee of $25 (£20) and $250 respectively. The M23 doubled the diggers’ wages to ensure they would carry on working.
It runs the area as a monopoly making sure – through the threat of arrest and detention – that only its authorised traders are able to do business.
The M23 also charges a levy of $7 on each kilogramme of coltan. The UN group of experts estimated that as a result the M23 earns about $800,000 a month from coltan taxation in Rubaya. That money is almost certainly then used to fund the rebellion.
There is a question mark hanging over how the ore extracted from M23-controlled areas gets into the global supply chain.
Neighbouring Rwanda, which is seen as backing the M23, is at the centre of the answer, the UN experts say.
Theoretically, a certification scheme – known as the Innovative Tin Supply Chain Initiative (Itsci) – should mean that what goes into a phone handset and other electronics does not come from areas of conflict where it could be used to fund armed groups responsible for carrying out atrocities.
The M23 is suspected of using the money raised in controlling the coltan mines to pay for its fighters and weapons
The US’ Dodd-Frank Act passed in 2010, and a similar piece of EU legislation, is aimed at ensuring that companies purchasing tin, tantalum, tungsten and gold – so-called “conflict minerals” – are not inadvertently funding violence.
But Itsci has come under some criticism.
Ken Matthysen, a security and resource management expert with independent research group Ipis, highlights that the dispersed nature of a lot of small-scale mines make it difficult for the local authorities to monitor exactly what is going on everywhere.
Itsci tags should be put on bags at the mine itself, to prove the origin of the minerals inside, but often they get transported to a collection point where it becomes harder to trace where the ore actually came from, Mr Matthysen said.
He added that there is also a possible issue with corruption.
“There is even an accusation of the state agents selling tags to traders, because they don’t make a good living. So the traders then go around eastern DR Congo and they tag the bags themselves.”
Itsci did not respond to a BBC request for comment, but has in the past defended its record saying that the scheme has been subjected to a rigorous independent audit. It has also been praised for bringing “prosperity for hundreds of thousands of small-scale miners”.
In the case of Rubaya, Itsci suspended its operations there soon after the M23 entered the town.
Nevertheless, the group has managed to continue exporting coltan.
The UN experts map a circuitous route showing how it is transported to close to the Rwandan border. It is then transferred to “heavy-duty trucks” that needed the road to be widened in order to accommodate them.
Rwanda has its own coltan mines but the experts say that the uncertified coltan is mixed with Rwandan production leading to a “significant contamination of supply chains”.
The M23 was already involved in the coltan business before the capture of Rubaya – setting up roadblocks and charging fees to cross them, according to Mr Matthysen.
“A lot of the trade of these minerals went through M23-controlled area towards Rwanda. So even then, Rwanda was profiting from the instability in eastern DR Congo and we saw the export volumes to Rwanda were already increasing,” he told the BBC.
The M23 increased the pay for the diggers in Rubaya but made sure they had a monopoly in the coltan trade
Figures from the US Geological Survey show that Rwanda’s coltan exports rose by 50% between 2022 and 2023. Mr Matthysen said this could not have all come from Rwanda.
In a robust defence of Rwanda’s position, government spokesperson Yolande Makolo reiterated to the BBC that there were minerals and refining capacity in her own country.
“It’s very cynical to take an issue like what’s happening in eastern DRC, where a persecuted community is fighting for its rights… and turning [it] into an issue of material benefit,” she added.
Rwandan President Paul Kagame has also dismissed the UN experts’ reports, pouring scorn on their “expertise”.
Much of the east of DR Congo has been blighted by conflict for many years, raising questions about who has been benefitting and whether armed groups are profiting from what is dug out of the ground there.
In order to highlight the issue and its connection to the smartphone industry, the Congolese government filed criminal complaints in France and Belgium at the end of last year against subsidiaries of the tech giant Apple, accusing it of using “conflict minerals”.
Apple has disputed the allegation and pointed out that since early 2024, because of the escalating conflict and the difficulties of certification, it stopped sourcing tantalum, among other metals, from both DR Congo and Rwanda.
Other companies have not been so clear, which means that as the M23 seizes more territory those small bits of tantalum from the mines that they control could still make their way into the devices that we have come to rely on.
Scores of journalists and civil society members were targeted on WhatsApp by Israeli spyware firm Paragon Solutions, according to a report published Friday.
In all, WhatsApp said it has “high confidence” that 90 journalists and members of civil society were targeted and “possibly compromised,” the company said in a statement to the British newspaper The Guardian.
It is unclear who directed the cyberattack, but Paragon’s Graphite spyware is known to be used by government clients.
Meta, WhatsApp’s parent company, did not immediately respond to a request for comment.
WhatsApp told the British daily that it issued a “cease and desist” letter to Paragon, notifying the Israeli company that it is weighing potential legal options in response to the hack. Notifications to victims remain ongoing.
“WhatsApp has disrupted a spyware campaign by Paragon that targeted a number of users, including journalists and members of civil society. We’ve reached out directly to people who we believe were affected. This is the latest example of why spyware companies must be held accountable for their unlawful actions. WhatsApp will continue to protect people’s ability to communicate privately,” a spokesperson said.
Once infected by Graphite, a user’s phone is completely compromised, and the spyware operator can access everything on it, including encrypted messages.
Anonymous experts told The Guardian that the hack was likely a “zero-click” attack, which does not require a user to click on any links to become infected.
WhatsApp said it believes users became infected by a malicious PDF file sent to targets who were added to group chats, saying it has assessed with “confidence” that Paragon was involved in the attack.
Meta has agreed to pay President Donald Trump $25 million to settle a 2021 lawsuit he filed claiming he was wrongfully censored by Facebook and Instagram after the US Capitol riot, the company said Wednesday.
The Wall Street Journal first reported the settlement of the suit brought against Meta and its chief executive Mark Zuckerberg, in what was seen as a victory for Trump.
According to people familiar with the agreement, the Journal said, $22 million of the payment will go towards funding Trump’s future presidential library, with the remainder covering legal fees and payments to other plaintiffs in the case.
Meta in the settlement will not admit wrongdoing over the suspensions of Trump’s accounts.
Trump had widely criticized social media platforms for suspending his accounts after the January 6, 2021 insurrection by his supporters, and comments he made that were seen as praising people engaged in the violence.
But he has recently courted tech titans including Zuckerberg and X owner Elon Musk, both of whom attended Trump’s presidential inauguration last week in Washington.
Zuckerberg has expressed support for Trump, and he has tweaked Meta’s policies to lift restrictions on some content within the company’s apps, which include Facebook, Instagram, Threads and WhatsApp.
Meta would be “restoring free expression on our platforms,” Zuckerberg, who reportedly dined with Trump at his Florida estate in November, said this month in announcing a rollback of fact-checking operations.
The settlement is the latest bow by media corporations as they gird for a second Trump presidency.
In December, ABC News agreed to pay a $15 million settlement payment to resolve a defamation lawsuit brought by Trump stemming from on-air comments about him made by a top anchor.
Earlier Wednesday Meta reported its net income soared by 59 percent to $62.36 billion for the full year.
Chinese tech company Alibaba on Wednesday released a new version of its Qwen 2.5 artificial intelligence model that it claimed surpassed the highly-acclaimed DeepSeek-V3.
The unusual timing of the Qwen 2.5-Max’s release, on the first day of the Lunar New Year when most Chinese people are off work and with their families, points to the pressure Chinese AI startup DeepSeek’s meteoric rise in the past three weeks has placed on not just overseas rivals, but also its domestic competition.
“Qwen 2.5-Max outperforms … almost across the board GPT-4o, DeepSeek-V3 and Llama-3.1-405B,” Alibaba’s cloud unit said in an announcement posted on its official WeChat account, referring to OpenAI and Meta’s most advanced open-source AI models.
The Jan. 10 release of DeepSeek’s AI assistant, powered by the DeepSeek-V3 model, as well as the Jan. 20 release of its R1 model, has shocked Silicon Valley and caused tech shares to plunge, with the Chinese startup’s purportedly low development and usage costs prompting investors to question huge spending plans by leading AI firms in the United States.
But DeepSeek’s success has also led to a scramble among its domestic competitors to upgrade their own AI models.
Two days after the release of DeepSeek-R1, TikTok owner ByteDance released an update to its flagship AI model, which it claimed outperformed Microsoft-backed OpenAI’s o1 in AIME, a benchmark test that measures how well AI models understand and respond to complex instructions.
This echoed DeepSeek’s claim that its R1 model rivalled OpenAI’s o1 on several performance benchmarks.
DEEPSEEK VERSUS DOMESTIC COMPETITORS
The predecessor of DeepSeek’s V3 model, DeepSeek-V2, triggered an AI model price war in China after it was released last May.
The fact that DeepSeek-V2 was open-source and unprecedentedly cheap, only 1 yuan ($0.14) per 1 million tokens – or units of data processed by the AI model – led to Alibaba’s cloud unit announcing price cuts of up to 97% on a range of models.
Other Chinese tech companies followed suit, including Baidu, which released China’s first equivalent to ChatGPT in March 2023, and the country’s most valuable internet company Tencent.
Liang Wenfeng, DeepSeek’s enigmatic founder, said in a rare interview with Chinese media outlet Waves in July that the startup “did not care” about price wars and that achieving AGI (artificial general intelligence) was its main goal.
OpenAI defines AGI as autonomous systems that surpass humans in most economically valuable tasks.
While large Chinese tech companies like Alibaba have hundreds of thousands of employees, DeepSeek operates like a research lab, staffed mainly by young graduates and doctorate students from top Chinese universities.
Liang said in his July interview that he believed China’s largest tech companies might not be well suited to the future of the AI industry, contrasting their high costs and top-down structures with DeepSeek’s lean operation and loose management style.
“Large foundational models require continued innovation, tech giants’ capabilities have their limits,” he said.
The emergence of Chinese AI app DeepSeek has shocked financial markets, and promptedUS President Donald Trump to describe it as “a wake-up call”for the US tech industry.
DeepSeek’s claim that its R1 artificial intelligence (AI) model was made at a fraction of the cost of its rivalshas raised questions about the future about of the whole industry, and caused some the world’s biggest companies to sink in value.
DeepSeek has become the most downloaded free app in the US just a week after it was launched.
So how does it compare to its much more established and apparently much more expensive US rivals, such as OpenAI’s ChatGPT and Google’s Gemini?
Writing Assistance
When you ask ChatGPT what the most popular reasons to use ChatGPT are, it says that assisting people to write is one of them.
From gathering and summarising information in a helpful format to even writing blog posts on a topic, ChatGPT has become an AI companion for many across different workplaces.
As a proud Scottish football fan, I asked ChatGPT and DeepSeek to summarise the best Scottish football players ever, before asking the chatbots to “draft a blog post summarising the best Scottish football players in history”.
DeepSeek responded in seconds, with a top ten list – Kenny Dalglish of Liverpool and Celtic was number one. It helpfully summarised which position the players played in, their clubs, and a brief list of their achievements.
DeepSeek also detailed two non-Scottish players – Rangers legend Brian Laudrup, who is Danish, and Celtic hero Henrik Larsson. For the latter, it added “although Swedish, Larsson is often included in discussions of Scottish football legends due to his impact at Celtic”.
For its subsequent blog post, it did go into detail of Laudrup’s nationality before giving a succinct account of the careers of the players.
ChatGPT’s answer to the same question contained many of the same names, with “King Kenny” once again at the top of the list.
Its detailed blog post briefly and accurately went into the careers of all the players.
It concluded: “While the game has changed over the decades, the impact of these Scottish greats remains timeless.” Indeed.
For this fun test, DeepSeek was certainly comparable to its best-known US competitor.
Coding
The emergence of advanced AI models has made a difference to people who code.
When ChatGPT experienced an outage last week, X had a number of amusing posts from developers saying they couldn’t do their work without the faithful tool by their side.
How does DeepSeek compare here?
Javier Aguirre, an AI researcher at Samsung Medical Center in Seoul, South Korea, specialises in researching in medicine and AI.
In a post on LinkedInon Tuesday, he wrote: “I am quite impressed with Deepseek. While coding, we usually try to explode AI chatbots to the limit to see their capabilities in assisting with coding.
“Today I had a really tricky and complex problem. Even chatGPT o1 was not able to reason enough to solve it. I gave a try to Deepseek and it solved it at once and straight to the point.”
He also pointed out that for coders, the combination of models can lead to success. This was echoed by Addy Osmani, who is the Head of Chrome Developer Experience at Google.
In a post to his208k followers on LinkedIn, he spoke about combining DeepSeek with US AI firm Anthropic’s tool Claude Sonnet.In 2023, Amazon invested $4bn into Anthropic.
Mr Osmani said: “Code with AI? DeepSeek R1 + Claude Sonnet may be the best new hybrid coding model. Yes, engineers are using them together.”
Mr Osmani also said DeepSeek was “significantly cheaper” to use than both Claude Sonnet and OpenAI’s o1 model.
Brainstorming ideas
What about brainstorming? I asked ChatGPT and DeepSeek to give me “ideas for a story for children about a boy who lives on the moon”.
ChatGPT responded in seconds with six neatly summarised ideas. One was about a boy called Max who worked as a postman on the moon and was sent on an adventure. Another was about Oliver, who was drawn by the sounds of a mysterious orchestra made up of aliens.
None of these stories are going to challenge Harry Potter or Roald Dahl any time soon, but it is a start for more refined ideas to flourish perhaps.
DeepSeek, on the other hand, responded with just one idea – “Luna and the Boy Who Chased the Stars”. Its response was 387 words (with no mention of anyone or thing called “Luna”), and comprised a story about a curious boy called Milo who lived on the moon.
It struck me that while ChatGPT gave me ideas, DeepSeek wrote a full story. It wasn’t particularly good, with a simple focus on a character going from A to B, but it was a start – and it was impressive how quickly it delivered it.
Learning and research
One of my memories from high school is my history teacher explaining to us how the First World War came about following a complex situation regarding many European powers, with the conflict finally sparked by the assassination of Archduke Franz Ferdinand in 1914.
How would the chatbots deal with explaining such a complex and nuanced piece of history? Pretty well.
ChatGPT gave a detailed account and outlined the key factors. DeepSeek’s account was not as detailed, but its brief overview did cover all the main points and events.
Google’s Gemini assistant gave a similar synopsis to ChatGPT and DeepSeek, and also gave the user the opportunity to click on links from reputable sources such as the Imperial War Museum in the UK.
As I saw on other tasks and prompts, DeepSeek was certainly comparable to its US competitors.
Steaming ahead
The tasks I set the chatbots were simple but they point to something much more significant – the winner of the so-called AI race is far from decided.
For all the vast resources US firms have poured into the tech, their Chinese rival has shown their achievements can be emulated.
Prof Neil Lawrence, DeepMind Professor of Machine Learning at Department of Computer Science and Technology, at the University of Cambridge, said this was just the start.
“I think it’s just the tip of the iceberg in terms of the type of innovation we can expect in these models,” he said.
He highlighted an example from history – James Watt is synonymous with the steam engine, even though he improved it rather than invented it.
“There’s plenty of space for budding James Watts to emerge, and that they are less likely to come from established players,” he said.
Apple’s iPhone devices are now eligible to test SpaceX-owned Starlink’s direct-to-cell capability that provides coverage from space, according to T-Mobile, a partner in the program.
T-Mobile and Elon Musk’s SpaceX are currently testing the Starlink cell network on a trial basis after receiving approval from the Federal Communications Commission in November last year.
The trial offers ‘text via satellite’, while voice and data features will be added in the future, according to the T-Mobile website.
T-Mobile initially only listed a few Android smartphones as eligible devices to test the network, but has now added iPhone devices with the latest iOS 18.3 software update.
Bloomberg News first reported on Tuesday that Apple, SpaceX and T-Mobile had been secretively working to add support for the network in its latest iPhone software.
Apple and SpaceX did not immediately respond to Reuters’ requests for comment outside regular business hours.
In October last year, the FCC had allowed SpaceX and T-Mobile to enable Starlink satellites with direct-to-cell capability to provide coverage for cellphones in areas of North Carolina hit hard by Hurricane Helene.
Safaricom is planning to build its own subsea cable to compete with Elon Musk’s Starlink.
This move comes as Safaricom seeks to enhance its internet services and maintain its market lead amid growing competition.
The subsea cable will provide high-speed internet and improve connectivity, reducing reliance on third-party providers.
These cables are the backbone of the global internet, carrying the bulk of international communications, including email, webpages and video calls Safaricom is seeking more bandwidth as it expands in the fast-growing data segment in the face of slower growth from the voice business.
According to reports, Safaricom is seeking approval from the Communications Authority of Kenya (CA) to build its own undersea cable as part of its strategy to expand and improve internet connectivity in the region.
Starlink has become an enormous competitor to Safaricom in the Kenyan market. Starlink, owned by Elon Musk, uses low-earth orbit satellites to deliver high-speed internet, making it an appealing solution for areas with limited access to traditional broadband infrastructure. Its ability to reach remote and underserved regions has created a niche market.
Currently, Safaricom relies on third parties for connections to the undersea cables.
This includes privately owned SEACOM and Telkom Kenya, which operates and maintains five of the six submarine cables that have landed in the country.
The five include the East African Marine Cable (TEAMS) cable, which is owned 32 percent by Safaricom, Telkom Kenya, 23 percent and the Kenyan government, 20 percent. It links Kenya to the outside world through the United Arab Emirates.
Others are EASSy, Lion 2, DARE 1 and PEACE subsea cable systems that are under Telkom Kenya.
Safaricom’s agreement with SEACOM is set to end in June 2028.
Analysts reckon that Safaricom is racing to diversify its sources of high-speed internet and cut reliance on Telkom Kenya in its quest for more bandwidth.
Starlink, which is riding on the back of one of the world’s richest persons with a net worth of $237 billion (Sh30.6 trillion), is betting on lowering internet costs in a segment dominated by Safaricom, Jamii Telecommunications Limited (JTL) and Zuku.
It has unsettled local telecoms players, with Safaricom, Airtel Kenya and Jamii Telecom having sent protest letters to the CA. The antitrust authority—Competition Authority of Kenya (CAK)—has also been dragged in court over Starlink’s operation.
The number of Kenyans using satellite internet has increased since Starlink entered the Kenyan market in July 2023.
US President Donald Trump has said that Microsoft is in discussions to acquire TikTok and that he would like to see a “bidding war” over the sale of the social media app.
When asked by reporters whether the US tech giant was preparing a bid, Trump replied: “I would say yes” – before adding that there was “great interest in TikTok” from several companies.
Both Trump and his predecessor Joe Biden have been trying for years to force TikTok’s Chinese parent company, ByteDance, to sell its US operations on national security grounds.
It comes as Trump signed an executive order last week to reverse a Biden Administration ban on TikTok that briefly took the app offline for its 170m users in the United States.
Despite granting TikTok a 75-day reprieve from the ban, Trump had been the first president to start pressuring ByteDance to sell its app.
In August 2020, ByteDance approached Microsoft as a possible buyer – something which the US company’s chief executive later described as “the strangest thing”.
Later, TikTok chose rival Oracle as a potential partner – although that deal also never happened.
Trump has previously said that he was in discussions with several parties about purchasing TikTok and expects to make a decision on the app’s future within the next 30 days.
A spokesperson for Microsoft said the company had “nothing to share at this time”. The BBC has also reached out to TikTok for comment.
Earlier on Monday, the US president had addressed a gathering of Republican politicians in Florida and spoke about the proposed sale of TikTok.
“We’ll see what happens. We’re going to have a lot of people bidding on it,” he said.
“If we can save all that voice and all the jobs, and China won’t be involved, we don’t want China involved, but we’ll see what happens,” he added.
Previous names linked with buying TikTok include billionaire Frank McCourt and the Canadian businessman Kevin O’Leary – a celebrity investor on Shark Tank, the US version of Dragon’s Den.
The biggest YouTuber in the world Jimmy Donaldson – AKA MrBeast – has also claimed he is in the running after a number of investors contacted him following an earlier tweet signalling his interest.
Chinese startup DeepSeek said on Monday it will temporarily limit registrations due to a cyberattack after the company’s AI assistant amassed sudden popularity.
The startup earlier in the day was also hit by outages on its website after its AI assistant became the top-rated free application available on Apple’s App Store in the United States.
The company resolved issues relating to its application programming interface and users’ inability to log in to the website, according to its status page. The outages on Monday were the company’s longest in around 90 days and coincides with its sky-rocketing popularity.
DeepSeek last week launched a free assistant it says uses less data at a fraction of the cost of incumbent players’ models, possibly marking a turning point in the level of investment needed for AI.
Powered by the DeepSeek-V3 model, which its creators say “tops the leaderboard among open-source models and rivals the most advanced closed-source models globally”, the artificial intelligence application has surged in popularity among U.S. users since it was released on Jan. 10, according to app data research firm Sensor Tower.
The milestone highlights how DeepSeek has left a deep impression on Silicon Valley, upending widely held views about U.S. primacy in AI and the effectiveness of Washington’s export controls targeting China’s advanced chip and AI capabilities.
Technology stocks were hammered on Monday, sending the shares of Nvidia and Oracle plummeting.
AI models from ChatGPT to DeepSeek require advanced chips to power their training. The Biden administration has since 2021 widened the scope of bans designed to stop these chips from being exported to China and used to train Chinese firms’ AI models.
However, DeepSeek researchers wrote in a paper last month that the DeepSeek-V3 used Nvidia’s H800 chips for training, spending less than $6 million.
Although this detail has since been disputed, the claim that the chips used were less powerful than the most advanced Nvidia products Washington has sought to keep out of China, as well as the relatively cheap training costs, has prompted U.S. tech executives to question the effectiveness of tech export controls.
Little is known about the company behind DeepSeek, a small Hangzhou-based startup founded in 2023, when search engine giant Baidu released the first Chinese AI large-language model.
Since then, dozens of Chinese tech companies large and small have released their own AI models, but DeepSeek is the first to be praised by the U.S. tech industry as matching or even surpassing the performance of cutting-edge U.S. models.
Chinese startup DeepSeek’s launch of its latest AI models, which it says are on a par or better than industry-leading models in the United States at a fraction of the cost, is threatening to upset the technology world order.
The company has attracted attention in global AI circles after writing in a paper last month that the training of DeepSeek-V3 required less than $6 million worth of computing power from Nvidia H800 chips.
DeepSeek’s AI Assistant, powered by DeepSeek-V3, has overtaken rival ChatGPT to become the top-rated free application available on Apple‘s App Store in the United States.
This has raised doubts about the reasoning behind some U.S. tech companies’ decision to pledge billions of dollars in AI investment and shares of several big tech players, including Nvidia, have been hit.
Below are some facts about the company shaking up the AI sector worldwide.
Why is DeepSeek causing a stir?
The release of OpenAI’s ChatGPT in late 2022 caused a scramble among Chinese tech firms, who rushed to create their own chatbots powered by artificial intelligence.
But after the release of the first Chinese ChatGPT equivalent, made by search engine giant Baidu, there was widespread disappointment in China at the gap in AI capabilities between U.S. and Chinese firms.
The quality and cost efficiency of DeepSeek’s models have flipped this narrative on its head. The two models that have been showered with praise by Silicon Valley executives and U.S. tech company engineers alike, DeepSeek-V3 and DeepSeek-R1, are on par with OpenAI and Meta‘s most advanced models, the Chinese startup has said.
They are also cheaper to use. The DeepSeek-R1, released last week, is 20 to 50 times cheaper to use than OpenAI o1 model, depending on the task, according to a post on DeepSeek’s official WeChat account.
But some have publicly expressed scepticism about DeepSeek’s success story.
Scale AI CEO Alexandr Wang said during an interview with CNBC on Thursday, without providing evidence, that DeepSeek has 50,000 Nvidia H100 chips, which he claimed would not be disclosed because that would violate Washington’s export controls that ban such advanced AI chips from being sold to Chinese companies. DeepSeek did not immediately respond to a request for comment on the allegation.
Bernstein analysts on Monday highlighted in a research note that DeepSeek’s total training costs for its V3 model were unknown but were much higher than the $5.58 million the startup said was used for computing power. The analysts also said the training costs of the equally-acclaimed R1 model were not disclosed.
Who is behind DeepSeek?
DeepSeek is a Hangzhou-based startup whose controlling shareholder is Liang Wenfeng, co-founder of quantitative hedge fund High-Flyer, based on Chinese corporate records.
Liang’s fund announced in March 2023 on its official WeChat account that it was “starting again”, going beyond trading to concentrate resources on creating a “new and independent research group, to explore the essence of “AGI” (Artificial General Intelligence). DeepSeek was created later that year.
ChatGPT makers OpenAI define AGI as autonomous systems that surpass humans in most economically valuable tasks.
It is unclear how much High-Flyer has invested in DeepSeek. High-Flyer has an office located in the same building as DeepSeek, and it also owns patents related to chip clusters used to train AI models, according to Chinese corporate records.
High-Flyer’s AI unit said on its official WeChat account in July 2022 that it owns and operates a cluster of 10,000 A100 chips.
How does Beijing view DeepSeek?
DeepSeek’s success has already been noticed in China’s top political circles. On January 20, the day DeepSeek-R1 was released to the public, founder Liang attended a closed-door symposium for businessman and experts hosted by Chinese premier Li Qiang, according to state news agency Xinhua.
Liang’s presence at the gathering is potentially a sign that DeepSeek’s success could be important to Beijing’s policy goal of overcoming Washington’s export controls and achieving self-sufficiency in strategic industries like AI.
A similar symposium last year was attended by Baidu CEO Robin Li.
Chinese startup DeepSeek’s AI Assistant on Monday overtook rival ChatGPT to become the top-rated free application available on Apple’s App Store in the United States.
Powered by the DeepSeek-V3 model, which its creators say “tops the leaderboard among open-source models and rivals the most advanced closed-source models globally”, the artificial intelligence application has surged in popularity among U.S. users since it was released on Jan. 10, according to app data research firm Sensor Tower.
The milestone highlights how DeepSeek has left a deep impression on Silicon Valley, upending widely held views about U.S. primacy in AI and the effectiveness of Washington’s export controls targeting China’s advanced chip and AI capabilities.
AI models from ChatGPT to DeepSeek require advanced chips to power their training. The Biden administration has since 2021 widened the scope of bans designed to stop these chips from being exported to China and used to train Chinese firms’ AI models.
However, DeepSeek researchers wrote in a paper last month that the DeepSeek-V3 used Nvidia’s H800 chips for training, spending less than $6 million.
Although this detail has since been disputed, the claim that the chips used were less powerful than the most advanced Nvidia products Washington has sought to keep out of China, as well as the relatively cheap training costs, has prompted U.S. tech executives to question the effectiveness of tech export controls.
Little is known about the company behind DeepSeek, a small Hangzhou-based startup founded in 2023, when search engine giant Baidu released the first Chinese AI large-language model.
Since then, dozens of Chinese tech companies large and small have released their own AI models, but DeepSeek is the first to be praised by the U.S. tech industry as matching or even surpassing the performance of cutting-edge U.S. models.
It offers a “PhD-level” AI at an economical rate of $2.19 per million output tokens compared to OpenAI’s $60 for similar usage. Some industry professionals have highlighted that this affordability has not been widely discussed, particularly in sell-side analyses. This omission could lead to uncertainties regarding the wider adoption and perception of DeepSeek’s capabilities.
The company has been commended for its use of reinforcement learning techniques, which reportedly optimize training costs and reduce complexity. DeepSeek claims its 1.5 billion-parameter R1 model outperforms industry standards, including GPT-4 and Claude 3.5, in select tasks, with minimal hardware requirements. For instance, the R1 model can reportedly operate on devices as simple as an iPhone 16.
Despite these breakthroughs, questions have been raised about DeepSeek’s originality. Critics have likened its approach to copying, citing statements from prominent figures like Sam Altman, who emphasized the risks of derivative models in achieving long-term success. Additionally, DeepSeek’s $6 million training budget has sparked skepticism regarding its scalability compared to competitors with larger investments.